United States
Environmental Protection
Agency
Oftice ol
Toxic Substances
Washington DC 20460
EPA-560/5-82-006
October, 1982
Toxic Substances
«>EPA
Analytical Methods
for By-Product PCBs-
Preliminary Validation
and Interim Methods
1000
1000
1000
ll..
'3C12H6CI4
280 285 290 295 300 305 310 315 320 325
Daltons
-------
DISCLAIMER
This document has been reviewed and approved for publication by the
Office of Toxic Substances, Office of Pesticides and Toxic Substances, U.S.
Environmental Protection Agency. Approval does not signify that the contents
necessarily reflect the views and policies of the Environmental Protection
Agency, nor does the mention of trade names or commercial products constitute
endorsement or recommendation for use.
-------
ANALYTICAL METHODS FOR BY-PRODUCTS PCBs—PRELIMINARY
VALIDATION AND INTERIM METHODS
By
Mitchell D. Erickson, John S. Stanley, Kay Turman, Gil Radolovich,
Karin Bauer, Jon Onstot, Donna Rose, and Margaret Wickham
Midwest Research Institute
^ 425 Volker Boulevard
Kansas City, MO 64110
TASK 51
INTERIM REPORT NO. 4
EPA Contract No. 68-01-5915
MRI Project.No. 4901-A(51)
October 11, 1982
For
U.S. Environmental Protection Agency
Office of Toxic Substances
Field Studies Branch
TS-798
Washington, D.C. 20460
Attn: Dr. Frederick W. Kutz, Project Officer
Mr. David P. Redford, Task Manager
-------
PREFACE
This report presents the results of a preliminary method validation ac-
complished on MRI Project No. 4901-A, Task 51, "PCB Analytical Methodology
Task," for the Environmental Protection Agency (EPA Prime Contract No. 68-
01-5915) during the period April 24 to August 31, 1982.
The document was prepared by Drs. Mitchell D. Erickson (Task Leader) and
John S. Stanley and Ms. Kay Turman, with assistance from Kathy Funk, Cindy
Melenson, and Gloria Sultanik. The laboratory work was conducted by Kay
Turman, and Donna Rose, with assistance from Steven Turner. The gas chro-
matography/mass spectrometry analysis was performed by Gil Radolovich,
Margaret Wickham, Jon Onstot, and Arbor Drinkwine. Statistical analysis of
the data was provided by Karin Bauer. Editorial comments were provided by
Rudena Mallory and Jeanne Robson.
The EPA Task Manager, David Redford, has been especially helpful and en-
couraging. The helpful comments of Ann Carey, Frederick W. Kutz, and John
Smith, all of EPA, are also appreciated.
Z SEARCH INSTITUTE
<-
n ^
E. Going, Head
Environmental Analysis Section
Approved:
James L. Spigarelli, Director
Analytical Chemistry Department
iii
-------
CONTENTS
Preface iii
Figures vii
Tables ix
1. Introduction 1
2. Summary 2
3. Experimental 3
Preparation of PCB stock solutions and working standards. 3
Gas chromatography/electron impact mass spectrometry. . . 8
Determination of PCB response factors (GC/EIMS) 8
Validation of method steps 19
Validation with product and product waste samples .... 20
4. Method Validation 24
Preparation of analytical methods 24
Gas chromatography/mass spectrometry of PCBs 26
Validation of selected method steps 38
Validation of product and product waste method with
samples 45
Discussion 56
5. References 70
' /
Appendix A - Supplementary GC/EIMS Data on PCB Congeners A-l
Appendix B - Analytical Method: The Analysis of Incidentally
Generated Chlorinated Biphenyls in Commercial Products and Product
Wastes B-l
Appendix C - Analytical Method: The Analysis .of Incidentally
Generated Chlorinated Biphenyls in Air ... C-l
Appendix D - Analytical Method: The Analysis of Incidentally
Generated Chlorinated Biphenyls in Industrial Wastewater D-l
-------
FIGURES
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
Plot of average response factor versus homolog for 77 PCB
congeners
Plot of response factor per isomer versus homolog for 77
PCB congeners
Plot of response factor per isomer versus homolog for 77
PCB congeners, determined on a single day
Retention times of 77 PCB congeners relative to 3,3*4,4'-
tetrachlorobiphenyl-d6 (RRT of 1.00)
Capillary gas chromatography/electron impact ionization
mass spectrometry (CGC/EIMS) chromatogram or the cali-
bration standard solution required for quantitation of
PCBs by homolog
Reconstructed ion chromatogram for SIM analysis of the
CMA-A sample no. 2110
SIM ion plots for monochlorobiphenyls (188 and 190
Daltons) and the 13C6-monochlorobiphenyl surrogate (194
Daltons) in CMA-A sample no. 2110
SIM ion plots for dichlorobiphenyls (222 and 224 Daltons)
in CMA-A sample no. 2110
SIM ion plots for trichlorobiphenyls (256 and 258 Daltons)
in CMA-A sample no. 2110
SIM ion plots for tetrachlorobiphenyls (290 and 292
Daltons), 3,3' ,4,4' -tetrachlorobiphenyl-d6 (298 Daltons) ,
and the 13C12-tetrachlorobiphenyl surrogate (304 Daltons)
in CMA-A sample no. 2110
SIM ion plots for pentachlorobiphenyls (326 and 328
Daltons) in CMA-A sample no. 2110
SIM ion plots of hexachlorobiphenyls (360 and 362 Daltons)
in CMA-A sample no. 2110
SIM ion plots of heptachlorobiphenyls (394 and 396 Daltons)
in CMA-A sample no. 2110
Page
27
28
32
36
39
59
60
61
62
63
64
65
66
Vll
-------
FIGURES (continued)
14 SIM ion plots of octachlorobiphenyls (428 and 430 Daltons)
and the 13Ci2~octachlorobiphenyl surrogate (442 Daltons)
in CMA-A sample no. 2110 67
15 SIM ion plots of nonachlorobiphenyl (464 and 466 Daltons)
in CMA-A sample no. 2110 68
16 SIM ion plots of decachlorobiphenyl (498 and 500 Daltons)
and the 13C12-decachlorbiphenyl (510 Daltons) in CMA-A
sample no. 2110 69
Vlll
-------
TABLES
Number Page
1 Numbering of PCB Congeners 5
2 Working Solutions for PCB Response Factors 6
3 Approximate Concentration of Individual PCB Congeners in
Dilute Working Standards 7
4 Concentrations of Congeners in PCB Calibration Standards
(ng/ml) 9
5 Composition of Surrogate Spiking Solution (SS100) Contain-
ing 13C-Labeled PCBs 10
6 Operating Parameters for Capillary Column Gas Chromato-
graphic System 11
7 DFTPP Key Ions and Ion Abundance Criteria for Quadrupole
Calibration 12
8 Operating Parameters for Quadrupole Mass Spectrometer
System 13
9 Operating Parameters for Magnetic Sector Mass Spectrometer
System 14
10 Characteristic Single lion Monitoring (SIM) Ions for PCBs . 15
11 Limited Mass Scanning (LMS) Ranges for PCBs 16
12 Characteristic Ions for 13C-Labeled PCB Surrogates 17
13 Pairings of Analyte, Calibration, and Surrogate Compounds . 18
14 Commercial Product and Product Waste Stream Samples
Received for Preliminary Method Validation Studies. ... 21
15 Preliminary Method Validation Samples 22
16 Comparison of Average Relative Response Factors (RRF) for
77 Commercially Available PCB Congeners Measured Over
Several Days as Four Replicates Each Versus Single Mea-
surements of All Congeners in a Single Day 30
IX
-------
TABLES (continued)
Number Page
17 Average Relative Response Factors (RRF) for PCB Congeners
in Solution 1 Measured as Replicates on a Single Day
and as Single Measurements for Day-to-Day Basis 31
18 Measured Average Response Factor (RRF) and Corresponding
Upper and Lower 95% Confidence Limits 34
19 Relative Response Factors Measured Versus 3,3',4,4'-Tetfa-
chlorobiphenyl-de by Electron Impact Mass Spectrometry
Quadrupole (Finnigan 4023) and Magnetic Sector (Varian
(MAT 311A) Instruments 35
20 Relative Retention Time (RRT) Ranges of PCB Homologs Versus
d6-3,3',4,4'-Tetrachlorobiphenyl 37
21 Recovery Data for Acid Cleanup 40
22 Recovery Data for Florisil Column Protocol Cleanup 41
23 Recovery Data for Florisil Slurry Protocol Cleanup 42
24 Recovery Data for KOH Protocol Cleanup. 43
25 Recovery Data for Alumina Protocol Cleanup 44
26 Uncorrected PCB Concentrations ((Jg/g) in CMA-A Samples. . . 46
27 Corrected PCB Concentrations (Mg/g) in CMA-A Samples. ... 47
28 Uncorrected and Corrected PCB Concentrations (Mg/g) in
CMA-E Sample (Dilution Preparation) 49
29 Uncorrected PCB Concentration (Mg/g) in the CMA^A Sample
Matrix (Internal Standard Calculation) 50
30 Corrected PCB Concentration (Mg/g) in the CMA-A Sample
Matrix 51
31 Uncorrected PCB Concentration (Mg/g) of Spiked CMA-A
Samples Determined by the Internal Standard Quantitation
Method 52
32 Corrected PCB Concentration (Mg/g) of Spiked CMA-A Samples
Determined by Surrogate Recovery Correction 53
33 PCB Concentration (Mg/g) of CMA-A Samples Heated With Dif-
ferent Cleanup Procedures (Internal Standard Quantita-
tion) 54
-------
TABLES (continued)
Number Page
34 PCB Concentration (|Jg/g) of CMA-A Samples Treated With
Various Cleanup Procedures (Surrogate Compound Cor-
receted) 55
35 Recovery (%) of Carbon-13 Labeled Surrogate Compounds
From Diarylide Yellow and Phthalocyanine Blue and Green
Pigments 57
XI
-------
SECTION 1
INTRODUCTION
The Environmental Protection Agency (EPA) is in the process of preparing
rules for regulation of certain polychlorinated biphenyls (PCB) which are
generated as by-products in the manufacture of commercial products (U.S. EPA,
1982). This regulation is under the Toxic Substances Control Act (PL 94-469),
and EPA's Office of Toxic Substances has been assigned the task of preparing
the rule.
As part of the rule, EPA is suggesting analytical methods for PCBs in
air (stack gas and fugitive emissions), wastewater, product waste streams,
and final products to assist organizations seeking an exclusion under this
rule. To assist EPA in this mission, Midwest Research Institute (MRI) was
asked to prepare appropriate analytical methodologies. A literature review
and recommendation of general analytical approaches (Erickson and Stanley,
1982; Stanley and Erickson, 1982) constituted the first phase. The second
phase, reported here, covers initial method validation and preparation of in-
terim methods. As part of the method validation, four 13C-PCB surrogates were
synthesized and are reported separately (Roth et al., 1982). The third phase
will involve interlaboratory validation and method refinement.
This report presents the initial results of method validation for analy-
sis of by-product PCBs in product and product waste samples. Specifically,
gas chromatography/electron impact mass spectrometry retention time and re-
sponse factor data for 77 PCB congeners for two different gas chromatography/
mass spectrometry systems, recoveries from several proposed cleanup steps,
and recoveries from industrial samples using a variety of the method options
are presented.
-------
SECTION 2
SUMMARY
The objective of this study was to present EPA with appropriate methodol-
ogies for the analysis of by-product PCBs in commercial products, product
waste streams, wastewaters, and air. In addition, EPA requested preliminary
analytical studies to provide data in support of the proposed methods.
This document presents proposed analytical methods for the analysis of
by-product polychlorinated biphenyls in commercial products and product waste
streams (Appendix B), wastewater (Appendix C), and air (Appendix D). The pro-
posed methods are based on determination of PCBs using gas chromatography/
electron impact mass spectrometry (GC/EIMS). Capillary column gas chroma-
tography (CGC) and packed column gas chromatography (PGC) are presented as
alternate approaches. The 13C-labeled PCS surrogates are added to samples
prior to any sample preparation to allow method flexibility for a wide spec-
trum of matrices. Recovery of the surrogates will allow determination of the
quality of analytical data. This method is valid only if the surrogates are
thoroughly incorporated into the matrix.
The analytical method for commercial products and product waste streams
relies heavily on a strong quality assurance program consisting of use of
four 13C-labeled surrogate PCBs, blanks, duplicates, spiked samples, and
quality control samples. The analytical methods for water and wastewater are
based on EPA Methods 608 and 625, revised to include the use of the 13C-
labeled surrogates. Likewise, the air method is a revision of a proposed
method for PCBs in air and flue gas emissions.
This document presents relative response factors (RRF) of 77 PCB congeners
which were used to determine the average RRF for PCBs by homolog. Statistical
analysis of the data was performed to check the validity of the response
factor data and to extrapolate RRFs for the unavailable congeners. Relative
retention time (RRT) data for the 77 PCB congeners are also presented. The
RRF and RRT data were determined on both magnetic sector and quadrupole mass
spectrometer systems.
Preliminary studies were undertaken to check the validity of the pro-
posed methods for the analysis of PCBs in commercial products and product
waste streams. Data are presented for analysis, of individual cleanup pro-
cedures as well as for analysis of product and product waste samples. The
data indicate that the proposed method is applicable and useful for analysis
of the matrices studied. However, these studies are preliminary and addi-
tional validation is necessary and ongoing.
-------
SECTION 3
EXPERIMENTAL
The method validation was conducted in three stages: (a) determination
of GC/EIMS parameters for 77 PCB congeners; (b) validation of individual
method steps with clean matrices; and (c) validation of selected method op-
tions with real samples.
PREPARATION OF PCB STOCK SOLUTIONS AND WORKING STANDARDS
Source of Standards
Seventy-seven PCB congeners were acquired from Ultra Scientific, Inc.,
Hope, Rhode Island, and Analabs, North Haven, Connecticut. Quality control
gas chromatography/flame ionization detection (GC/FID) data for the specific
isomers were requested to verify the 99% purity assigned to these compounds.
The GC/FID data supported the reported purity. In addition, all available
nuclear magnetic resonance spectra used for specific isomer identification
were requested but not supplied.
Weighing Procedures
Accurate mass measurement required calibration of a Cahn microbalance
with National Bureau of Standards (NBS) certified masses of 5 and 10 mg. The
balance was calibrated with the NBS standards followed by calibration of an
in-house working standard mass. The calibration of the microbalance with the
NBS certified masses was witnessed by a representative of the MRI quality as-
surance office. The mass of the working standard was measured between all
measurements of individual PCB isomers to ensure that the balance was operat-
ing accurately. A record of the measured working standard mass was kept in a
laboratory notebook. The mean value for the working standard was 10.037 ±
0.002 mg (0.02% relative standard deviation). When all measurements were com-
pleted, the mass of the NBS certified standards was determined as a final mea-
sure of the accuracy of the Cahn microbalance.
Preparation of Solutions
Preparation of PCB standard stocks began after accurate performance of
the Cahn balance was demonstrated with the certified NBS and daily working
standard. An aluminum weighing pan was preshaped such that complete transfer
of the weighing pan plus sample could be made directly into the appropriate
dilution vessel. The Cahn balance was tared to compensate for the weight of
the aluminum boat, and the PCB standards were added via a micro spatula. The
mass of the particular PCB was determined with the Cahn balance.
-------
The aluminum pan containing the PCB standard was transferred to the di-
lution vessel using clean forceps, taking care not to spill any of the sample.
The dilution vessel was capped tightly until solvent was added.
All PCB congeners were dissolved in toluene (Burdick and Jackson, dis-
tilled in glass). Masses of 0.1 to 5 mg were dissolved in a total of 1.0 ml
toluene while masses of approximately 10 mg and greater were dissolved in
5.0 ml toluene. The solvent was delivered volumetrically by pipette. Room
temperature and solvent temperature were recorded at the time of standard
dissolution. Volumetric pipettes used for solvent delivery were calibrated
so that the most accurate determination of analyte concentration could be
calculated. Toluene was pipetted into a tared vessel, and the total mass was
measured. Density of the solvent at the specific room temperature was used
to calculate the actual volume dispensed. This calibration was performed for
all pipettes used for volumetric delivery of solvent. The stock solutions
were sonicated in an ultrasonic bath for at least 15 sec after the volumetric
addition of toluene to ensure complete dissolution of the PCBs. The solution
level was etched on the side of the dilution vessel as a means of detecting
losses by evaporation.
The individual PCB congeners were referred to by the congener number in-
dicated in Table 1. The stable labeled PCBs, 3,3',4,4'-tetrachlorobiphenyl-d6,
4-chlorobiphenyl-13C«, 3,3',4,4'-tetrachlorobiphenyl-13C12, 2,2',3,3',5,5' ,6,6,''
octachlorobiphenyl-13C12, and decachlorobiphenyl-^C^ were assigned congener
numbers of 210 to 214, respectively, for the purpose of this work. Sample
labels were generated in duplicate to identify the specific PCB isomer stock
solution and to document entries in the laboratory notebook. Table 2 presents
the dilute working solutions that were prepared for determination of the re-
sponse factors for the PCB congeners. The working solutions were prepared as
10 ml total volume. Table 3 presents the approximate concentration of each
congener that was in the dilute working standard used for response factor de-
termination. Tetrachlorobiphenyl-dg was added to 1.0 ml of each solution as
the internal standard. All stocks were added to the working solutions in vol-
umes of 20, 200, 250, 400, 500, or 1,000 pi. The syringes were calibrated at
these volumes. Calibration of the 10-ml volumetric flasks used for working
standards was accomplished by measuring the difference between the mass of
the empty flask and the mass of the flask plus toluene added to the appropri-
ate dilution mark. The density of toluene at the correct solvent temperature
was used to calculate the final volume of each solution.
The dilute working solutions were divided into multiple aliquots. One
hundred micrograms of tetrachlorobiphenyl-d6 was added to each of the 1.0-ml
aliquots of the solutions that were used to establish CGC/EIMS response factors.
The remaining dilute working solutions were stored in at least four crimp seal
vials and refrigerated. The solvent meniscus was marked in permanent form to
note losses of solvents from evaporation or spills. All solutions, stock
standards and working solutions, were stored in a refrigerator. All vials
removed from storage were first brought to room temperature and then sonicated
for at least 15 to 30 sec before removing any of the solution.
-------
TABLE 1. CUMBERING OF PCB CONGENERS3
tJo.
1
2
3
4
5
5
7
3
9
10
11
12
13
14'
15
16
17
13
19
20
22
23
24
25
26
27
23
29
30
31
32
33
34
25
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Structure
HonadilorobiBhCTylj
2
3
4
Q1eh1arBblBh«ny1i
2.2'
2.3
2.3'
2.4
2,4'
2,5
2,6
3.3'
3,4
3.4'
3,5
4.4'
THeMorob
-------
TABLE 2. WORKING SOLUTIONS FOR PCB RESPONSE FACTORS
PCB
homolog
Monochloro-
Dichloro-
Trichloro-
Tetrachloro-
Pentachloro-
Hexachloro-
Heptachloro-
Octachloro-
Nonachloro-
Decachloro-
Total
congeners
Soln.
no. 1
1
11
29
47
121
136
181
195
207
209
10
Soln.
no. 2
2
5
21
44
97
129
171
194
208
9
PCB congener no.
Soln. Soln. Soln. Soln. Soln. Soln. Soln. Soln. Soln. Soln. Soln. Soln.
no. 3 no. 4 no. 5 no. 6 no. 7 no. 8 no. 9 no. 10 no. 11 no. 12 no. 13 no. 14
3
7 8 9 10 4 12 14 15
31 26 24 28 18 33 30
40 49 50 52 53 54 66 61 65 69 72 70,75,77
88 93 101 103 100 104 a 115 87 116 119
128 137 138 141 143 151 139 153 154 155 156
183 185
198 200 202 204
206
97665544333 3
a Congener no. 112 was added to this solution but, on analysis, was determined to have a mass of 286 and appeared
to be a diaminotrichlorobiphenyl. This congener was omitted from any further consideration.
-------
TABLE 3. APPROXIMATE CONCENTRATION OF INDIVIDUAL PCB CONGENERS
IN DILUTE WORKING STANDARDS3
PCB homolog Concentration (pg/ral)
Monochlorobiphenyl 50
Dichlorobiphenyl 50
Trichlorobiphenyl 50
Tetrachlorobiphenyl 100
Pentachlorobiphenyl 100
Hexachlorobiphenyl 100
Heptachlorobiphenyl 100
Octachlorobiphenyl 200
Nonachlorobiphenyl 200
Decachlorobiphenyl 200
a Tetrachlorobiphenyl-de was added to all solutions as an internal standard
at ~ 100 M8/ml-
-------
Preparation of Calibration Standard and Spiking Mixtures
A mixture of 11 congeners was used for calibration. This solution was
spiked into solvent for protocol step validation experiments and into product
and product waste samples for standard addition experiments. These congeners
were determined to be the best standards for quantitation calibration based
on the average relative response factor for each PCB homolog, as will be dis-
cussed in Section 5.
Table 4 presents the composition of the 11-component solutions that are
specified as the calibration standards, CSxxx, where the xxx is used to en-
code the nominal concentration in nanograms per milliliter. A more con-
centrated solution was diluted as necessary to prepare spiked samples and
the appropriate standards for GC/EIMS analysis. The internal standard, tetra-
chlorobiphenyl-dg, was added to all standards and final extracts before GC/
EIMS analysis. The standards contained the four 13C-labeled PCBs that were
added from the spiking solution shown in Table 5.
GAS CHROMATOGRAPHY/ELECTRON IMPACT MASS SPECTROMETRY
The capillary gas chromatography parameters used are shown in Table 6.
The quadrupole and magnetic sector mass spectrometer parameters used are
shown in Tables 7 through 9. The characteristic ions for single ion monitor-
ing and limited mass scanning are presented in Tables 10 through 12.
All data generated for relative response factors and concentration levels
of PCBs in sample extracts were calculated based on the area of the primary
quantitation ion specified in Table 10. The quantitation ions for the 13C-
labeled monochloro-, tetrachloro-, octachloro-, and decachlorobiphenyl were
194, 304, 442, and 510 Daltons, respectively. The pairings of analyte, cali-
bration, and surrogate compounds are presented in Table 13.
DETERMINATION OF PCB RESPONSE FACTORS (GC/EIMS)
The response factors for 77 PCB isomers were determined by GC/EIMS using
the working standards prepared as described in Tables 2 and 3. A high reso-
lution capillary column (J&W Scientific Durabond DB-5, 15 m, 0.25 |Jm film
thickness) was used for the separation of the PCB mixtures. Scanning mass
spectrometry was used to calculate response factors for the PCB isomers
present in each solution versus a known quantity of tetrachlorobiphenyl-d6.
The quadrupole GC/EIMS system was tuned daily prior to any acquisition
of data for PCB response factor calculations. The system was brought to op-
erating temperature for at least 15 rain. The fluorocarbon FC-43 was intro-
duced to the ion source, and 176 and 502 Daltons were manually adjusted to a
two-to-one ratio. This was accomplished by adjusting the multiplier voltage
to 300 mV while monitoring 176 Daltons. A selected ion monitor acquisition
was set up for 176 and 502 Daltons with a variance of 1 Dalton. The ratio of
the two values was tuned to the two-to-one ratio as described above. The mass
spectrometer was operated in the normal full scan acquisition mode after tun-
ing with the FC-43. Approximately 100 ng of decafluorotriphenylphosphine was
injected and the ratio of the values of 198/442 was monitored.
-------
TABLE 4. CONCENTRATIONS OF CONGENERS IN PCB CALIBRATION STANDARDS (ng/ml)a
Homolog
1
1
2
3
4
5
6
7
8
9
10
4
1
4
8
10
Congener
no.
1
3
7
30
50
97
143
183
202
207
209
210 (IS)
211 (RS)
212 (RS)
213 (RS)
214 (RS)
CS1000
1,040
1,000
1,040
1,040
1,520
1,740
1,920
2,600
4,640
5,060
4,240
255
104
257
407
502
CS100
104
100
104
104
152
174
192
260
464
506
424
255
104
257
407
502
CS050
52
50
52
52
76
87
96
130
232
253
212
255
104
257
407
502
CS010
10
10
10
10
15
17
19
26
46
51
42
255
104
257
407
502
a Concentrations given as examples only.
-------
TABLE 5. COMPOSITION OF SURROGATE SPIKING SOLUTION (SS100)
CONTAINING 13C-LABELED PCBsa
Congener
no.
211
212
213
214
Concentration
Compound (pg/ml)
(I1, 2', 3
(13C12)3
(13C12)2
1 ,4' ,5',6'-13C6)4-chlorobiphenyl
,3' ,4,4' -tetrachlorobiphenyl
, 2 ' , 3, 3 ' , 5, 5 ' , 6, 6 ' -octachlorobiphenyl
(13C12)decachlorobiphenyl
104
257
395
502
a Concentrations given as examples only.
10
-------
TABLE 6. OPERATING PARAMETERS FOR CAPILLARY COLUMN GAS CHROMATOGRAPHIC SYSTEM
Parameter
Value
Gas chromatograph
Column
Liquid phase
Liquid phase thickness
Carrier gas
Carrier gas velocity
Injector
Injector temperature
Injection volume
Initial column temperature
Column temperature program
Separator
Transfer line temperature
Finnigan 9610
15 m x 0.255 mm ID
Fused silica
DB-5 (J&W)
0.25 |Jm
Helium
45 cm/sec
On-column (J&W)
Optimum performance
i.o Mib
110°C (2 min)C
. d
110° to 325°C at 10°C/min
None
280°C
a Measured by injection of air or methane at 270°C oven temperature.
b For on-column injection, follow J&W instructions regarding injection tech-
nique.
c With on-column injection, the initial temperature equals the boiling point
of the solvent; in this instance toluene.
d C12Cl1o elutes at 270°C. Programming above this temperature ensures a
clean column and lower background on subsequent runs.
e Fused silica columns may be routed directly into the ion source to pre-
vent separator discrimination and losses.
11
-------
TABLE 7. DFTPP KEY IONS AND ION ABUNDANCE
CRITERIA FOR QUADRUPOLE CALIBRATION
Mass Ion abundance criteria
197 Less than 1% of mass 198
198 100% relative abundance
199 5-9% of mass 198
275 10-30% of mass 198
365 Greater than 1% of mass 198
441 Present but less than mass 443
442 Greater than 40% of mass 198
443 17-23% of mass 442
12
-------
TABLE 8. OPERATING PARAMETERS FOR QUADRUPOLE MASS SPECTROMETER SYSTEM
Parameter Value
Mass spectrometer Finnigan 4023
Data system Incos 2400
Scan range 95-550
Scan time 1 sec
Resolution Unit
Ion source temperature 280°C
Electron energy 70 eV
Trap current 0.2 mA
Multiplier voltage -1,600 V
Preamplier sensitivity 10 6 A/V
a Filaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no separator is used.
13
-------
TABLE 9. OPERATING PARAMETERS FOR MAGNETIC SECTOR MASS SPECTROMETER SYSTEM
Parameter Value
Mass spectrometer Finnigan MAT 311A
Data system Incos 2400
Scan range 98-550
Scan mode Exponential
Cycle time 1.2 sec
Resolution 1,000
Ion source temperature 280°C
o
Electron energy 70 eV
Emission current 1-2 mA
Filament current Optimum
Multiplier -1,600 V
a Filaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no separator is used.
14
-------
TABLE 10. CHARACTERISTIC SINGLE ION MONITORING (SIM) IONS FOR PCBs
Homolog
Ci2H9Cl
Ci2H8Cl2
C12H7C13
cl2Hecl4
C12H5C15
C12H4C16
C12H3C17
Ci2H2Clg
C12HClg
Cl2CllO
Primary
188 (100)
222 (100)
256 (100)
292 (100)
326 (100)
360 (100)
394 (100)
430 (100)
464 (100)
498 (100)
Ion (relative intensity)
Secondary
190 (33)
224 (66)
258 (99)
290 (76)
328 (66)
362 (82)
396 (98)
432 (66)
466 (76)
500 (87)
Tertiary
a
226 (11)
260 (33)
294 (49)
324 (61)
364 (36)
398 (54)
428 (87)
462 (76)
496 (68)
Source: Rote JW, Morris WJ. 1973. Use of isotopic abundance ratios in
identification of polychlorinated biphenyls by mass spectrometry.
J Assoc Offie Anal Chem 56(1):188-199.
/
a None available.
15
-------
TABLE 11. LIMITED MASS SCANNING (IMS) RANGES FOR PCBs
Compound
C12H9C11
Ci2HgCl2
C12H7C13
Ci2H6Cl4
C12H5C15
C12H4C16
C12H3C17
Ci2H2Clg
C12HC19
C12CllO
C12D6C14
13C612C6H9C1
13C12H6C14
I
13C12H2C18
13Ci2Clio
r*
Mass range (Daltons)
186-190
220-226
254-260
288-294
322-328
356-364
386-400
426-434
460-468
494-504
294-300
192-196
300-306
438-446
506-516
a Adapted from Tindall GW, Wininger PE. 1980. Gas chromatography-mass
spectrometry method for identifying and determining polychlorinated
biphenyls. J Chromatogr 196:109-119.
16
-------
TABLE 12. CHARACTERISTIC IONS FOR 13C-LABELED PCB SURROGATES
Compound
13C612C6H9C1
13C12H6C14
13C12H2C18
13Ci2Cl10
Primary
194 (100)
304 (100)
442 (100)
510 (100)
Ion (relative intensity)
Secondary
196 (33)
306 (49)
444 (65)
512 (87)
Tertiary
a
302 (78)
440 (89)
514 (50)
a None available.
17
-------
TABLE 13. PAIRINGS OF ANALYTE, CALIBRATION, AND SURROGATE COMPOUNDS
Analyte
Congener
no.
1
2,3
4-15
16-39
40-81
82-127
128-169
170-193
194-205
206-208
209
Compound
Calibration standard
Congener
no . Compound
2-C12H9Cl 1
3- and 4-C12H9Cl 3
Ci2HgCl2 7
Ci2H?Cl3
C12H6C14
C12H5C15
C12H4C16
C12H3C17
Ci2H2Cls
Cj^HClg
Ci2dio
30
50
97
143
183
202
207
209
2
4
2,4
2,4,6
2, 2', 4, 6
2, 2', 3', 4, 5
2, 2', 3,4, 5, 6'
2, 2', 3', 4, 4', 5', 6
2,2', 3,3', 5, 5', 6, 6'
2, 2', 3, 3', 4, 4' ,5, 6, 6'
C12Clio
Congener
no.
211
211
211
212
212
212
212
213
213
213
214
Surrogate
Compound
13c6-4
13C6-4
13C6-4
13Ci2-3,3'
13Ci2-3,3'
13C12-3,3'
13Ci2-3,3'
13C12-2,2'
13C12-2,2'
13C12-2,2'
13Ci2Cl10
,4
,4
,4
,4
,3
,3
,3
,4'
,4'
,4'
,4'
,3'
,3'
,3'
,5
,5
,5
,5'
,5'
,5'
,6
,6
,6
,6'
,6'
,6'
00
-------
The response of 198 Daltons was 100% full scale and 442 Daltons was adjusted
from 40 to 45% of the base peak. These criteria were met daily before data
acquisition for response factor calculations was initiated.
All working standards were brought to room temperature and sonicated be-
fore injection into the GC/MS system. Solution No. 1 was analyzed daily as a
means of normalizing response factors calculated from day to day. This al-
lowed some compensation for differences in sensitivity due to subtle changes
in the mass spectrometer operation from day to day. Also, a solution of tetra-
chlorobiphenyl-de (internal standard) was analyzed separately. Four replicates
of each working standard were analyzed to calculate variances of the response
factors. The solutions were sonicated at least 15 sec prior to removal of
sample for injection. The syringe and needle were rinsed with 200- to 300-|Jl
of toluene between injections.
The gas chromatograph was operated at 110°C for 2 min, and programmed at
10°C/min to 325°C. One microliter injections were made with a J&W on-column
injection system. Helium carrier flow was adjusted to 45 cm/sec.
The peak shape of the eluting PCBs was monitored. If excessive tailing
was noted, the injection end of the fused silica capillary column was removed
and shortened by at least 10 cm.
Tables 6, 7, and 8 present the instrument and operating parameters that
were used to measure the response factors for the individual PCB isomers in
the working solutions. Response factors (RF) were calculated using the area
of the peaks for these ions according to the equation:
RTr _ APCB MIS
~ A M
Ais "PCB
A
where JPCB = Area of the quantitation peak of the specific PCB,
.IS = Mass (in nanograms) of the internal standard injected,
MIS = Area of the quantitation peak of the internal standard, and
PCB = Mass (nanograms) of the specific PCB injected.
All relative response factor data were subjected to Student's t-test at
the 95% confidence level to test for significant differences for day-to-day
and solution-to-solution variances.
VALIDATION OF METHOD STEPS
A limited number of experiments were completed as preliminary validation
steps for the proposed method presented in Appendices B through D. The ex-
periment included evaluation of several of the cleanup procedures using solvent
spiked with the 13C-labeled surrogates and a mixture of PCB congeners repre-
senting each of the possible homologs. The laboratory cleanup procedures fol-
lowed the protocol steps except where noted. One hexane solvent blank was
analyzed by each procedure with the samples to monitor interferences and con-
tamination.
19
-------
All samples were analyzed by CGC/EIMS in the full scan mode using the
Finnigan 4023 system. Tables 6, 7, and 8 present the instrumental parameters.
VALIDATION WITH PRODUCT AND PRODUCT WASTE SAMPLES
Sources of Samples
Product waste samples were received from Dow Chemical Company (Kent Hodges)
and Vulcan Materials Company (Thomas Robinson) through the cooperation of the
Chemical Manufacturers Association (Robert Fensterheim). These samples are
aliquots of the materials used for the Chemical Manufacturers Association
(CMA) round robin study (CMA, 1982). The CMA and associates supplied samples
of chlorinated benzene waste streams, mixtures of chlorinated benzenes, com-
posite waste streams from a chlorinated aliphatic process and a benzene column
bottom sample. Table 14 presents an inventory of all the samples received.
Product samples were received from the Dry Color Manufacturers Associa-
tion (J. Lawrence Robinson and Maria DaRoche). These samples included diarylide
yellow, phthalocyanine green, and phthalocyanine blue pigments that were used
in the Dry Color Manufacturers Association (DCMA) round robin study of an an-
alytical method, reported by the DCMA (1981). These samples are also included
in the inventory in Table 14.
The samples supplied by industry are examples of the samples which will
be analyzed using the method in Appendix B. However, since no attempt was
made to span the range of products and product wastes, the samples analyzed
do not include all matrices which an analyst could encounter.
Experimental Design
Table 15 presents an overview of the preliminary method validation sam-
ples. The samples from Table 14 that were used for these studies included
the chlorinated benzene waste stream, CMA-A; the benzene column bottom sample,
CMA-E; and the yellow, blue, and green pigment samples, DCMA-1, DCMA-4, and
DCMA-8, respectively. Blind quantitation standards and quality control sam-
ples were prepared by the MRI quality control staff either through spiked ad-
dition or by dilution of particular sample matrices. Other quality control
procedures included the analysis of duplicate samples and blanks and the
validation of cleanup steps. Two sets of samples were prepared and run at
separate times. This first sample set us designated by numbers 10 through
110 and the second sample set is designated by numbers 2001 through 2210Q in
Table 15.
The sample preparations ranged from addition of the 13C-labeled sur-
rogates followed by dilution and injection, to preparation of pigment samples
via sulfuric acid dissolution and hexane extraction or methylene chloride ex-
traction with Florisil cleanup.
20
-------
TABLE 14. COMMERCIAL PRODUCT AND PRODUCT WASTE STREAM SAMPLES
RECEIVED FOR PRELIMINARY METHOD VALIDATION STUDIES3
Sample no. Quantity
Sample description
Sample source
CMA-A
CMA-B
CMA-C
CMA-A
CMA-B
CMA-C
CMA-D
CMA-E
100 ml Chlorinated benzene waste
stream
100 ml Mixture of chlorinated benzenes
with Aroclor 1254 spike
100 ml Blind spike of CMA-B with the
addition of 64 ppm of PCB
isomers
5 ml Chlorinated benzene waste
stream
5 ml Mixture of chlorinated benzenes
with Aroclor 1254 spike
5 ml Blind spike of CMA-B with the
addition of 64 ppm of PCB
isomers
5 ml Composite waste stream sample
from a chlorinated aliphatic
process
5 ml Benzene column bottoms sample
Dow Chemical Co.
Dow Chemical Co.
Dow Chemical Co.
Vulcan Materials Co.
Vulcan Materials Co.
Vulcan Materials Co.
Vulcan Materials Co.
Vulcan Materials Co.
DCMA-1
DCMA-4
DCMA-6
DCMA-8
DCMA-9
100 g
100 g
100 g
100 g
100 g
Diarylide yellow pigment
Phthalocyanine green pigment
Phthalocyanine blue pigment
Phthalocyanine blue pigment
Phthalocyanine green pigment
DCMA
DCMA
DCMA
DCMA
DCMA
a Aliquots of CMA-A, CMA-B, and CMA-C were received from two sources, who
indicated that they were identical. MRI has assumed that both aliquots
are the same.
21
-------
TABLE 15. PRELIMINARY METHOD VALIDATION SAMPLES
Sample
no.
Description
Preparation
Dilution
factor
10 CMA-A
20A CMA-A
20B CMA-A
60 Hexane blank
110 CMA-E
2001 Hexane blank
2005 CMA-A3
2010 CMA-A
2020 CMA-A
2025Q CMA-A
2030 CMA-A + CS002
2040 CMA-A + CS005
2050 CMA-A + CS010,
2060Q CMA-A + CSXXX
2070Q CSxxx
2080 Blank,
2090 CMA-Ab
2100 Blank
2110 CMA-A
2120 Blank
2130 CMA-A
2135 DCMA-13
2140 DCMA-1
2150 DCMA-1
2160 DCMA-1
2170Q DCMA-1
2175 DCMA-4
2180 DCMA-4
2185 DCMA-4'
2190 DCMA-8
2195 DCMA-8J
2200Q DCMA-8
2210Q CSxxx
+ no. 11 (50 ppra)
+ no. 11 (20-80 ppm)
0.1 g/10 ml hexane 1/100
0.1 g/10 ml hexane 1/100
0.1 g/10 ml hexane 1/100
None None
None None
None None
0.1 g/1 ml hexane 1/10
0.1 g/1 ml hexane 1/10
0.1 g/1 ml hexane 1/10
0.5-0.2 g/1 ml hexane *> 1/10
0.1 g/1 ml hexane 1/10
0.1 g/1 ml hexane 1/10
0.1 g/1 ml hexane 1/10
0.1 g/1 ml hexane 1/10
None None
DCMA-A 1/10
DCMA-A (0.1 g) 1/10
DCMA-B 1/10
DCMA-B (0.1 g) 1/10
Base 1/10
Base (0.1 g) 1/10
DCMA-B (1.0 g) 1/100
DCMA-B (1.0 g) 1/100
DCMA-B (1.0 g) 1/100
DCMA-B (1.0 g) 1/200
DCMA-B (1.0 g) 1/200
DCMA-B 1/100
DCMA-B 1/100
DCMA-B 1/100
DCMA-A 1/50
DCMA-A 1/50
DCMA-A 1/50
None None
a No surrogates added to assess any background interferences for these
compounds.
b Prepared from aliquot received from Dow Chemical Company; all other CMA-A
samples prepared from aliquot received from Vulcan Materials Company.
22
-------
The CMA-A and CMA-E samples were each analyzed after 1/10 or 1/100 dilu-
tion, depending on the operating sensitivity of the mass spectrometer. The
CMA-A chlorinated benzene waste was the most extensively studied matrix of
the available samples. Sample preparation included the simple dilution de-
scribed above with and without the addition of the four surrogates. The sam-
ples prepared without surrogates allowed measurement of the background that
might interfere with the four surrogate compounds. Duplicate samples of the
CMA-A were analyzed at the same dilution in two separate experiments. The
CMA-A matrix was also analyzed by standard addition methods with total spiked
PCB levels of the 11-compound spiking solution (CS050) at approximately 70,
140, and 270 ng/sample. The CMA-A matrix was also prepared using the sul-
furic acid and ethanolic KOH procedures discussed in Section 9.3.2 of Ap-
pendix D, Cleanup of the Analytical Method: The Analysis of By-Product
Chlorinated Biphenyls in Commercial Product and Product Wastes (Appendix B).
Variations of the analytical procedures used by the Dry Color Manufacturers
Association (1981) for the analysis of PCBs in various pigments were also ap-
plied to the CMA-A matrix. The DCMA procedures included acid dissolution fol-
lowed by hexane extraction from the acid (DCMA Preparation A) and Florisil
treatment of the concentrated sample matrix (DCMA Preparation B). The homogen-
ization and centrifugation steps required by the DCMA-B procedure were not
included for the CMA-A matrix. All samples except those representing blanks
were spiked with the surrogates at levels of 100 to 500 ng and were mixed
thoroughly before beginning the sample preparation. The typical CMA-A sample
size was 0.1 g.
The diarylide yellow (DCMA-1), phthalocyanine green (DCMA-4), and pthalo-
cyanine blue (DCMA-8) pigments were also studied in these preliminary valida-
tions. The yellow pigment was prepared according to the recommended DCMA-B
procedure, while the green and blue pigments were analyzed following the DCMA-A
procedure. The preparation of the pigments followed the DCMA procedures except
that the preparation was scaled to 1 g of the yellow pigment instead of the
recommended 5 g. Blanks, duplicates, and spiked samples were also analyzed
with the set of DCMA samples.
Sample Analysis
All extracts were analyzed by capillary column gas chromatography/electron
impact mass spectrometry (CGC/EIMS). Limited mass scanning (LMS) or selected
ion monitoring (SIM) mass spectrometry methods were used for extract analy-
sis, depending on the level of PCBs in the sample extracts and the complexity
of the matrix. The parameters for analysis via CGC/LMS and CGC/ MS-SIM are
presented in Tables 6 through 13.
23
-------
SECTION 4
METHOD VALIDATION
PREPARATION OF ANALYTICAL METHODS
Analytical methods were prepared for the analysis of by-product PCBs in:
* Commercial products and product wastes (Appendix B).
"' Air (Appendix C).
* Industrial wastewater (Appendix D).
The analysis of commercial products and product wastes was covered in
one method since the diversity of matrices in both categories dictates the
same generalized approach. Air was defined to include stack gases, fugitive
emissions, and static (room, other container, or outside) air.
Commercial Products and Product Wastes Method
The objective was to devise an analytical method suitable for enforce-
ment of the regulation concerning by-product PCBs in commercial products and
product wastes. A detailed rationale for selection of the techniques used in
the method may be found in a separate report (Erickson and Stanley, 1982).
Sample Workup--
The general approach taken with sample preparation (collection, preser-
vation, extraction, and cleanup) was to provide a framework within which any
reasonable technique could be used. This is the only acceptable approach to
a method designed to cover "any" matrix. ,
The use of 13C-labeled recovery surrogates in conjunction with GC/EIMS
was judged to be the most suitable approach (Erickson and Stanley, 1982;
Stanley and Erickson, 1982; Roth et al., 1982). Using the recovery surrogates,
any losses of PCBs would be detected and could be corrected for in the calcu-
lation of the PCB concentration.
24
-------
When surrogates are not fully incorporated into the matrix, their re-
covery will not be representative of the analyte PCB recoveries and recovery
assessment will not be possible. It is incumbent upon the analyst to recog-
nize this problem and use good scientific judgment with samples that present
a potential problem. Nonextractable solid polymers may be an example of a
matrix presenting incorporation problems.
PCB Determination--
As discussed elsewhere (Erickson and Stanley, 1982; Stanley and Erickson,
1982), GC/EIMS appears to be the only acceptable general technique for deter-
mining PCBs in commercial products and product wastes. The use of either
capillary or packed column GC is permitted. While strong arguments are pre-
sented for both techniques (Stanley and Erickson, 1982), the analytical results
should be comparable for both techniques provided proper instrument calibra-
tion and operation, analytical, and quality control procedures are followed
as described in the analytical methods.
Quantitation--
The analytical objective of these methods is to determine if the sample
contains quantifiable PCBs and, if so, at what concentration. On the assump-
tion that a general knowledge of the congener distribution is important, re-
porting of the concentration by homolog is proposed in the reporting form.
Since a "total PCB" value is also important for summary and comparative pur-
poses, space for this value is also provided on the reporting form. Other
reporting formats, including "largest isomer or resolvable peak" or "all peaks
greater than a regulatory value," may be easily accommodated using different
tabulations and reporting procedures.
The PCB concentrations found may be lower than the actual value due to
nonquantitative recovery during extraction or cleanup. The measured recov-
eries of the surrogates may be used to derive a corrected concentration. The
analyst must take care that the surrogates are thoroughly incorporated into
the matrix prior to extraction, as discussed above. The analyst must also
guard against improper corrections because of errors in surrogate quantita-
tion. These errors may arise from background interferences. A more thorough .
discussion of quantitation options is presented in a previous report (Erickson
and Stanley, 1982).
Air Method
The sample collection, preservation, extraction, and cleanup aspects were
taken from the work of Haile and Baladi (1977). The determination, using GC/
EIMS, is identical to that in the commercial products and product wastes
method except that recovery surrogates are not used.
Wastewater Method I
The water method is a direct modification of the commercial products and
product wastes method. As noted in this method, the cleanup and extraction
procedures for EPA Methods 608 (U.S. EPA, 1979b) and 625 (U.S. EPA, 1979a)
may be used. It is anticipated that, unless conditions dictate otherwise,
most analysts will choose this option.
25
-------
Quality Control
Each method includes a strong quality control (QC) section. Given the
complexity of the matrices and complexity of the analyte (209 compounds), the
need for QC is evident. The various aspects of the QC section were designed
assuming a reasonably large (10 to 100) batch of samples. For small batches
of samples, the percentage of effort spent on QC can become sizeable.
Alternate Methods
The methods presented here are intended to be primary methods capable of
generating the best quality data technologically feasible. The development
and acceptability of secondary (alternate, equivalent, or screening) methods
is not addressed in this report.
GAS CHROMATOGRAPHY/MASS SPECTROMETRY OF PCBs
Analysis for PCBs requires the use of selected representative standard
compounds since all 209 congeners are not available. One of the major dis-
advantages of many instrumental methods for PCB analysis is the large vari-
ance of the instrumental response factors for PCB congeners, both within a
homolog and between homologs. These large differences in response factors
create problems in selecting representative compounds for quantitation pur-
poses. The response factors of 77 of the possible 209 PCB congeners measured
by GC/EIMS are presented in Tables 16 and 17. The data suggests that the EIMS
response factor variance among PCB congeners is small relative to other de-
tectors such as the electron capture detector or negative chemical ionization
mass spectrometry.
Relative Response Factors
Quadrupole Mass Spectrometer--
The relative response factors (RRF) of the 77 PCB congeners were deter-
mined with the Finnigan 4023 quadrupole mass spectrometer as discussed in the
experimental section. The RRFs were determined two ways to assess the effects
of instrumental variability. The replicate RRF determinations are the average
of four replicate analyses for each of the PCB congeners, all determined on a
single day to assess the variability of the measurement. The single RRF de-
terminations are single values from an experiment in which all 14 solutions
containing all 77 congeners were run on one day to minimize instrumental vari-
ability with time. The data are presented in Appendix A. The RRFs vary from
approximately 0.2 for decachlorobiphenyl to 4.1 for 2-chlorobiphenyl. Fig-
ures 1 and 2 present a visual comparison of average replicate and single RRFs
of PCB congeners determined as replicate measurements and as single measure-
ments .
26
-------
4.5,—
4.0
3.5
o
J-l
o
-------
3.Or— 1
NJ
00
2.5
2.0
U
a)
U-i
SI1-5
C
o:
a
co-
ai
1.0
0.5
_L
2
3
2
1
3
1
Quadrupole Mass Spectrometer
_L
J_
34 5 6 7
Homolog (degree of chlorination)
10
Figure 2. Plot of response factor per isomer versus homolog for 77 PCB congeners, determined on
a single day. Each value is representative of single measurements of each congener with the
Finnigan 4023 quadrupole mass spectrometer. This plot indicates the number of data points that
overlap for specific isomers.
-------
Table 16 is a summary of the KRF data, where the replicate and single
measurements are averaged over all measured isomers for a homolog. The rela-
tive standard deviation (Table 16) for the replicate measurements reflects
the variance of the average RRF for each isomer within a homolog. The abso-
lute area of the internal standard, Congener No. 210, varied by only 4.4% for
all solutions during the single day experiment, as compared to 9.9% for the 7
days required to complete the replicate analyses. The relative standard devi-
ations based on the four replicate analyses for each of the PCS congeners,
ranged from 0.4 to 9.1%, indicating the reproducibility of the injection for
each solution.
The average response factors from replicate determinations and single
measurements were subjected to a Student's t-test to determine if there were
any significant differences in measured response factors. No significant
difference was found for the average response factor values for any of the
PCS homologs except the heptachlorobiphenyl isomers. A more detailed presen-
tation of the Student's t-test for these values is presented in Table A-2 of
Appendix A.
A solution of 3,3',4,4'-tetrachlorobiphenyl-de (Congener No. 210) and
Solution No. 1 (Table 2) were both analyzed daily. The solution of Isomer
No. 210 was used to tune the quadrupole mass spectrometer to the desired
working conditions. Solution No. 1 was used to determine fluctuations of re-
sponse factors from day to day due to differences in instrumental operating
parameters. Table 17 presents the data for single day replicate measurements
and day-to-day determination of the response factors for the PCB congeners in
Solution No. 1. The relative standard deviations calculated for the single
day measurements are considerably lower than the relative standard deviations
from day-to-day analyses. This is a reflection of the reproducibility on the
part of the operator as well as of the stability of the quadrapole mass spec-
trometer system on a given day. The relative standard deviation calculated
for day-to-day analyses is indicative of the variation that might be expected
for routine analysis of PCBs.
A Student's t-test of the Solution No. 1 data (Table 17) indicated that
there are significant differences in response factors from day to day compared
to single day measurements for PCB Congener Nos. 1, 11, 29, and 207. A more
detailed presentation of this t-test is presented in Table A-3 of Appendix A.
Magnetic Sector Mass Spectrometer--
The RRFs for the 77 PCB congeners were also determined with a Varian MAT
311A double focusing magnetic sector mass spectrometer. The RRF values were
determined by single measurements of all congeners on a single day. The data
are presented in Appendix A and summarized in Figure 3.
Extrapolation of Response Factor Data to All Congeners--
Since all 209 PCB congeners were not available for determination of RRFs,
it was necessary to extrapolate the average RRF data to project the range of
response factors that might be encountered. This extrapolation was based on
the assumption that the number of measured isomers (n) are a representative
sample of the entire set of the possible isomers (N). Thus it was assumed
that the mean for the measured isomers (n) is an unbiased estimate of the
mean for the possible isomers (N).
29
-------
TABLE 16. AVERAGE RELATIVE RESPONSE FACTORS (RRF) FOR 77 COMMERCIALLY AVAILABLE
PCB CONGENERS MEASURED OVER SEVERAL DAYS AS FOUR REPLICATES EACH AND RRF
FOR SINGLE MEASUREMENTS OF ALL CONGENERS IN A SINGLE DAY
PCB homolog
Monochloro-
Dichloro-
Trichloro-
Tetrachloro-
Pentachloro-
Hexachloro-
Heptachloro-
Octachloro-
Nonachloro-
Decachloro-
No. of
isomers
3
10
9
16
12
13
4
6
3
1
RRF from
replicate
^ a
measurements
3.331
2.027
1.573
0.950
0.720
0.513
0.361
0.253
0.229
0.213
Relative
standard
deviation (%)
19.3
22.0
21.7
18.4
16.7
15.1
6.6
11.9
14.7
2.8
RRF from
single ,
measurement
2.739
2.048
1.592
0.946
0.725
0.500
0.308
0.224
0.188
0.179
Relative
standard
deviation (%)
9.3
15.7
18.1
20.0
17.6
19.1
8.0
17.3
16.2
-
a Four replicate measurements of the RRF were made for each isomer. For example,
the three monochlorobiphenyl isomers were measured four times each. Hence,
the RRF and relative standard deviation (%) were calculated from 12 distinct
values.
b A single measurement for each of the 77 PCB congeners was completed in a single
day. Hence, the RRF reported is the average of one measured RRF for each
isomer within a homolog. For example, the RRF and relative standard deviation
(%) reported for the monochlorobiphenyls were calculated from three distinct
values.
30
-------
TABLE 17. AVERAGE RELATIVE RESPONSE FACTORS (RRF) FOR PCB CONGENERS IN
SOLUTION 1 MEASURED AS REPLICATES ON A SINGLE DAY AND AS
SINGLE MEASUREMENTS FOR DAY-TO-DAY BASIS3
Single day measurements
Congener
no.
1
11
29
47
121
136
181
195
207
209
RRF
4.073
3.073
2.195
1.062
0.948
0.689
0.383
0.263
0.237
0.213
Std.
deviation
0.118
0.073
0.048
0.059
0.020
0.016
0.009
0.003
0.008
0.006
Relative std.
deviation (%)
2.905
2.363
2.188
5.591
2.127
2.336
2.379
1.184
3.547
2.837
Day-to-day measurements
RRF
3.544
2.733
2.005
1.032
0.955
0.685
0.377
0.270
0.257
0.223
Std.
deviation
0.452
0.300
0.171
0.061
0.036
0.046
0.028
0.022
0.030
0.023
Relative std.
deviation (%)
12.767
10.977
8.535
5.876
3.747
6.688
7.347
8.304
11.757
10.352
a See Tables 6 and 8 for CGC/EIMS operating conditions.
b These values calculated from four replicates.
c These values calculated from 11 separate analyses.
31
-------
2.50
2.00
1.50
0)
CO
c
o
ex
W
0)
1.00
0.50
Magnetic Sector Mass Spectrometer
4567
Homolog (degree of chlorination)
10
Figure 3. Plot of response factor per isomer versus homolog for 77 PCB congeners, determined on a single
day. Each value is representative of single measurements of each congener with the Varian Mat 311A
magnetic sector mass spectrometer. This plot indicates the number of data points that overlap for
specific isomers.
-------
Table 18 presents the upper and lower 95% confidence limits for the mea-
sured average RRFs. The extrapolation was necessary for the dichloro- through
octachlorobiphenyl homologs. The projected upper and lower limits of the av-
erage RRF ranged from 13% for each PCB homolog for trichlorobiphenyls to ap-
proximately 6.5% for the dichlorobiphenyls. The projected ranges for the
tetrachloro- to octachlorobiphenyls were between these values.
Comparison of Magnetic Sector and Quadrupole RRF Data--
The two instruments used operate on entirely different principles, so
the results may represent the range of RRFs to be expected from these com-
pounds on different instruments. Table 19 presents a summary of the data.
As expected, the RRF trends are much different. Since quadrupole spectrom-
eters discriminate at the high masses, the RRFs for high homologs (higer
masses) are much lower than corresponding values for the magnetic detector
spectrometer.
A statistical analysis of the data (Student's t-test presented in Table 4
of Appendix A) confirmed that the average RRFs are significantly different
for many of the homologs. However, the relative standard deviations for the
average RRF of each homolog are not significantly different. Thus, the ex-
trapolation from a single calibration isomer to all isomers of a homolog should
have similar precision for the two instrument types.
Relative Retention Times
Relative retention times (RRT) were also calculated from the data gene-
rated for relative response factor measurements with both the quadrupole and
magnetic sector mass spectrometer instruments. All RRTs for each PCB congener
were calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6. Figure 4 is a plot
of the RRT data versus PCB homolog. All data points for the 77 PCB congeners
measured with the quadrupole mass spectrometer are presented. This plot also
indicates that the relative retention window for the dichloro- to octachloro-
biphenyl homologs may be larger than that actually measured if more of the
possible congeners were present.
Table 20 presents the observed range of RRTs for the 77 PCB congeners
and additional congeners, identified only by homolog, in an Aroclor mixture
(1016, 1254, 1260). These RRTs were established using a 15-m fused silica
DB-5 capillary column. It must be recognized that the RRT windows on other
columns may be substantially different. Table 20 also presents a projected
RRT window for PCB anaysis. The overlap of the retention windows of each
homolog must be considered in establishing an instrumental analysis approach
to quantitation of the specific PCB homologs. This consideration has been
accounted for in the GC/MS requirements for PCB analysis in Appendices B to
D. The relative retention times of the 77 PCB congeners as determined with
both the quadrupole and magnetic sector mass spectrometers are presented in
tabular form in Appendix A.
33
-------
TABLE 18. MEASURED AVERAGE RELATIVE RESPONSE FACTOR (RRF) AND
CORRESPONDING UPPER AND LOWER 95% CONFIDENCE LIMITS
PCB homolog
Monochloro-
Dichloro-
Trichloro-
Tetrachloro-
Pentachloro-
Hexachloro-
Heptachloro-
Octachloro-
Nonachloro-
Decachloro-
No. of
possible
isomers
(N)
3
12
24
42
46
42
24
12
3
1
No. of
available
isomers
(n)
3
10
9
16
12
13
4
6
3
1
Average
measured
response
RRF
3.331
2.027
1.573
0.950
0.720
0.513
0.361
0.253
0.229
0.213
Sample std. ,
deviation Lower Upper
(S) limit limit
0.643
0.447 1.896 2.158
0.341 1.366 1.780
0.175 0.877 1.023
0.120 0.654 0.786
0.078 0.474 0.552
0.024 0.326 0.396
0.030 0.231 0.275
0.034
- - -
4-_ / _\ 1.
a Lower 95% limit = RRF - — /1 - £
b Upper 95% limit = RRF + — 1 - £
V~ I rl
n V
34
-------
TABLE 19. RELATIVE RESPONSE FACTORS MEASURED VERSUS 3,3',4,4'-TETRACHLORO
BIPHENYL-d6 BY ELECTRON IMPACT MASS SPECTROMETRY QUADRUPOLE (FINNIGAN
4023) AND MAGNETIC SECTOR (VARIAN MAT 311A) INSTRUMENTS
PCB homolog
Monochloro-
Dichloro-
Trichloro-
Tetrachloro-
Pentachloro-
Hexachloro-
Heptachloro-
Octachloro-
Nonachloro-
Decachloro-
No. of
isomers
measured
3
10
9
16
12
13
4
6
3
1
RRF
Quadrupole_
Mean
2.739
2.048
1.592
0.946
0.725
0.500
0.308
0.224
0.188
0.179
RSD" (%)
9.3
15.7
18.1
20.0
17.6
19.1
8.0
17.3
16.2' -
-
Magnetic
Mean
2.329
1.663
1.167
0.902
0.780
0.640
0.497
0.463
0.467
0.586
sector
RSDa (%)
8.5
13.8
21.3
14.0
17.4
19.4
12.1
15.3
22.5
-
a Relative standard deviation.
35
-------
CO
C12H,CI9
C]2H2CI8
C,2H3CI7
S C12H4d6
o
o
X
pq
O
PM
C12H5CI5
Cl2H6CI4
C,2H7CI3
C,2H8CI2
C,2H9CI,
Relotive Retention Times of PCB Congeners by Homolog
Versus 3,3',4,4' Tetrachlorobiphenyl-d0
208*0' **
202200 lOfl 195 194
185 171
IB!) 181
IS4 139 153 138 129
155 136 151 143 141 137 128 156
» «««»«»»««i4 •-
.„. 121 9/IIS
104 103ICO 9388 "" 119 t15
-• •» «• • •«•--
75
«,, 6949
54 5° 53 5247 .
• » « ••—
70
'61 66 65 77
»•« • «-
2631 33
30 18 24 29 28 21
10 9 8 ,
4 7 5 14 1l"l5
23
J_
_L
_L
_L
0.40 0.50 0.60 0.70 0.80 0.90 1.00
Relative retention time
1.10
1.20
1.30
1.40
Figure 4. Retention times of 77 PCB congeners relative to 3,3',4,4'-tetrachlorobiphenyl-d5 (RRT of 1.00)
The dashed line indicates that not all of the possible isomers of a particular homolog were measured.
Relative retention times were determined on a J&W DB-5, 15-m fused silica column in a Finnigan 4023
GC/EIMS system. Temperature program: 110°C for 2 minr then 10°C/min to 325°C.
-------
TABLE 20. RELATIVE RETENTION TIME (RRT) RANGES OF PCB HOMOLOGS
VERSUS d6-3.3',4,4'-TETRACHLQROBIPHENYL
PCB
homolog
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
No. of
isomers
measured
3
10
9
16
12
13
4
6
3
1
Observed range
of RRTsa
0.40-0.50
0.52-0.69
0.62-0.79
0.72-1.01
0.82-1.08
0.93-1.20
1.09-1.31
1.19-1.36
1.31-1.42
1.44-1.45
Calibration
Congener
no.
1
3
7
30
50
97
143
183
202
207
209
solution
Observed
RRT3
0.43
0.50
0.58
0.65
0.75
0.98
1.05
1.15
1.19
1.33
1.44
Projected
range of
RRTs
0.35-0.55
0.35-0.80
0.35-1.10
0.55-1.05
0.80-1.10
0.90-1.25
1.05-1.35
1.10-1.50
1.25-1.50
1.35-1.50
a The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were
measured versus de-3,3',4,4'-tetrachlorobiphenyl (internal standard)
using a 15-m J&W DB-5 fused silica column with a temperature program of
110°C for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec,
and an on-column injector. A Finnigan 4023 Incos quadrupole mass spec-
trometer operating with a scan range of 95-550 Daltons was used to de-
tect each PCB congener.
b The projected relative retention windows account for overlap of eluting
homologs and take into consideration differences in operating systems
and lack of all possible 209 PCB congeners.
37
-------
Selection of Congeners for a Calibration Standard
The data generated from the RRF and RRT measurements were used to select
the PCB congeners for an analytical quantitation/calibration standard for
GC/EIMS analysis of PCBs. Selection of the standard compounds was based pri-
marily on the ratio of the measured response factor to the average response
factor for a particular homolog. The PCBs with RRFs closest to the average
values were selected as standard compounds. In addition, the RRT was con-
sidered to assure that the selected PCB congeners did not coelute. Two mono-
chlorobiphenyls were selected for the calibration standard because the aver-
age RRF and RRT did not clearly coincide with any of the three possible
isomers. One isomer (2-chlorobiphenyl) had a substantially different RRF.
This isomer was quantitated separately. 4-Chlorobiphenyl was selected as the
calibration isomer for the two remaining isomers. Figure 5 is a CGC/EIMS
chromatogram of the 11-component PCB calibration standard. The composition
of this solution is identified in Tables 4 and 20 along with the observed RRT
of each of the 11 congeners.
VALIDATION OF SELECTED CLEANUP STEPS
As part of the overall method validation, several of the cleanup tech-
niques were validated. A mixture of the 11 calibration standard congeners
and three recovery surrogates (the 13C-octachlorobiphenyl was unavailable for
these experiments) was diluted in an appropriate solvent and then subjected
to the cleanup procedures as described in Appendix B. After the cleanup, the
internal standard was added and the volume adjusted. The samples were analyzed
by CGC/EIMS using a quadrupole spectrometer operated under the condition listed
in Tables 6 through 8. Data were collected in the full scan mode and quanti-
tated using the primary ions listed in Table 10 and the congener pairs listed
in Table 13. A blank was run through the procedure alongside the recovery
spikes. As expected, no PCBs except the internal standard were observed in
the blanks.
The results for the 11 calibration congeners were calculated as percent-
age recovery. Tables 21 through 25 present the uncorrected recoveries, cal-
culated using Equation 12-1 of Appendix B, using the internal standard (Con-
gener No. 210); the actual percentage recoveries of the 13C-labeled recovery
surrogates, calculated using Equation 12-2 of Appendix B; and the corrected
recoveries of the calibration congeners, calculated using Equation 12-3 of
Appendix B.
Inspection of Tables 21 to 25 reveals that the accuracy of the corrected
recoveries is higher than for the uncorrected recoveries (104% versus 77%
average). On the other hand, the precision of the uncorrected recoveries is
slightly higher than for the corrected recoveries (Il7o versus 9% relative stan-
dard deviation average). This is the expected trend since the uncorrected
recovery relies on two GC/MS measurements (area of the PCB congener peak and
area of the internal standard peak) and the corrected recovery relies on those
two values and the area of the surrogate peak. Thus, these results indicate
that accuracy is improved by recovery correction, at a sacrifice of precision.
38
-------
100.0-1
me
u>
rH
r*1*
c
CU
XI
CX
•H
XI
0
M
0
rH
XI
o
O
C
•8
CN
6I8
|
523 575 H
GK3
10:0:)
VO
rH
1
C C
O> 01
x; x; -
ex ex
XI XI ^
O O 0)
M M -U
o o cd
rH rH 60
XI XI O
CJ O l-i
0 0 rl
C C 3
O O CO
S S ^
i I
•*
c
rH
C
01
-. XI
CX
0) -rl
rH
C
0)
XI
CX
XI
0
o
rH
XI
O
cd
XI XI 4J
C
•r-
i. O
J-l
XI 0
c
1-
c
1—
rH
XI
a
•H
XI rl
0)
H
vO
v^-
•.
CN
OH
*f
C
i
vO
CN CN
j
I.I
0 798
«
..,L82L
r
C'JO
13:20
CN
07'1
.
936
rH Cd
cd oo
C 0
O> t-l
4-1 3
C to
v~' u
vO CO
'^ rH
1 1
rH rH
r^"> r**»
C C
0) 0)
XI XI
ex ex
•H -rH
f O ^Q
0 0
)->
•U C
OJ OJ
•H rC
1 CX
•H
*3" Xi
- O
^J- 1-1
- o
- rH
CO XI
« CJ
ro cd
^
o>
SB
1
vO
"
"1
XI
CX
•rl
O
0
rH
O
cd
CX
cu
1
vO
M
»•
m
VI
-
"*,
^t
f^>
1101
,__(
^
rH C
>i O>
c x:
01 CX
XI -H rH
ex xt >->
•rl 0 C
Xi M CiJ
0 0 XI
1-1 rH CX
O X! -H
rH 0 XI
xi cd o
a C rl
cd o o
4J £7^ rH
O 1 X!
0 CJ
I vo cd
r, O
vO vO 0)
" « Q
vo m
•. »
in ^d"
^ n
in ^
•v n
•. to
CO CO
v> n
CO CO
«i *>
f •.
CN CN
CN CN
^ 1208 I3IG
CN
1178
Jfl
CN
0
CO
,__j
1
rH
'C
0>
XI
ex
•H
x>
0
0
rH
XI
o
cd
a
o>
o
,— s
cu
4J
cd
60
o
l-i
1-1
3
to
N-X
i ' l
2f>:&) 23:20
112592.
SCA:I
TIlli:
Figure 5. Capillary gas chromatography/electron impact ionization mass spectrometry (CGC/EIMS)
chromatogram or the calibration standard solution required for quantitation of PCBs by homolog,
This chromatogram includes PCBs representative of each homolog, three ^-^c-labeled surrogates,
.and the deliberated internal .standard. The concentration of all components and the CGC/EIMS
parameters are presented in Tables 4, 5, 6 and 9.
-------
TABLE 21. RECOVERY DATA FOR ACID CLEANUP'
Congener no .
1
3
7
30
50
97
143
183
202
207
209
X
Standard deviation
Relative standard deviation (%)
211
212
214
X
Standard deviation
Relative standard deviation (%)
Total spike
PCB homolog level (|Jg)
Monochlorobiphenyl
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nona chlo rob iphenyl
Decachlorobiphenyl
13Ce-nionochlorobiphenyl
13Ci2~tetrachlorobiphenyl
13C 12 -decachlorob iphenyl
0.52
0.50
0.52
0.52
0.76
0.87
0.96
1.30
2.30
2.50
2.10
2.60
5.30
10.20
Spike 2 (%
Uncorrected
100.0
83.4
82.5
NQ
78.0
99.5
81.2
85.9
80.7
83.2
87.3
86.2
7.5
9
70.2
87.1
91.3
82.9
11.2
13
recovery)
Corrected
142.4
118.8
117.5
NQ
89.6
114.2
93.2
98.5
88.4
91.1
95.7
104.9
17.7
17
-
-
-
Blank
NDC
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
-
-
a Spike No. 1 not analyzed.
b Corrected via surrogate response.
c Not detected.
d Large background signal prevented quantitation of the compound.
e Not applicable.
-------
TABLE 22. RECOVERY DATA FOR FLORISIL COLUMN CLEANUP
Congener no.
1
3
7
30
50
97
143
183
202
207
209
X
Standard deviation
.P, Relative standard
M deviation (%)
211
212
214
X
Standard deviation
Relative standard
deviation (%)
PCB homolog
Monochlorobiphenyl
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
13C6-monochlorobiphenyl
13C12-tetrachlorobiphenyl
13C12-decachlorobiphenyl
Total spike
level (fjg)
0.52
0.50
0.52
0.52
0.76
0.87
0.96
1.30
2.30
2.50
2.10
2.60
5.30
10.20
Spike 1 (%
Uncorrected
57.9
63.0
66.0
69.4
70.7
73.4
72.6
76.6
77.8
78.1
77.7
71.2
6.7
9
63.9
43.2
75.9
61.0
16.5
27
recovery)
Corrected
90.6
98.6
103.2
160.5
163.6
169.7
168.1
117.2
102.5
102.9
102.4
130.9
35.8
27
-
-
-
-
Spike 2 (%
Uncorrected
54.9
58.3
60.0
62.3
62.4
66.1
67.0
72.3
72.3
70.5
72.8
65.4
6.2
10
57.6
47.9
69.6
58.4
10.9
19
recovery)
Corrected
95.4
101.2
104.4
130.0
130.3
138.1
140.1
151.0
103.8
101.3
104.5
118.2
19.8
17
-
-
-
-
Blank
NDb
ND
ND
ND
ND
ND
ND
ND
ND
ND
NDc
-
-
-
ND
ND
ND
-
-
-
a Corrected via surrogate response.
b Not detected.
c Not applicable.
-------
TABLE 23. RECOVERY DATA FOR FLORISIL SLURRY CLEANUP
Congener no.
1
3
7
30
50
97
143
183
202
207
209
X
Standard deviation
Relative standard
deviation (%)
211
212
214
X
Standard deviation
Relative standard
deviation (%)
PCB homolog
Mpnochlorobiphenyl
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pent a chlo r ob ipheny 1
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nona chlorob ipheny 1
Decachlorobiphenyl
13C6-monochlorobiphenyl
13C12-tetrachlorobiphenyl
13C12-decachlorobiphenyl
Total spike
level ((jg)
0.52
0.50
0.52
0.52
0.76
0.87
0.96
1.30
2.30
2.50
2.10
2.60
5.30
10.20
Spike 1 (%
Uncorrected
80.5
81.2
87.5
NQC
90.0
96.0
95.5
95.1
97.2
95.1
96.2
91.4
6.3
7
83.9
98.3
106.5
92.5
7.5
8
recovery)
Corrected"
96.0
96.8
104.4
NQ
91.6
97.6
97.2
96.8
91.2
89.4
90.4
95.1
4.5
5
_
-
-
-
-
-
Spike 2 (%
Uncorrected
71.1
72.7
75.0
76.4
80.1
83.5
82.0
79.8
88.8
87.6
83.7
80.1
5.8
7
76.7
89.4
87.9
84.7
6.9
8
recovery)
Corrected
92.9
94.8
98.1
85.5
89.6
93.5
91.6
89.3
101.0
99.5
95.2
93.7
4.7
5
_
-
-
-
-
-
Blank
NDb
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND,
Q
-
-
ND
ND
ND
-
-
-
a Corrected via surrogate response.
b Not detected.
c Large background signal prevented quantitation of this compound.
d Not applicable.
-------
TABLE 24. RECOVERY DATA FOR KOH CLEANUP
CO
Congener no.
1
3
7
30
50
97
143
183
202
207
209
X
Standard deviation
Relative standard
deviation (%)
211
212
214
X
Standard deviation
Relative standard
deviation (%)
PCB homolog
Monochlorobiphenyl
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
13Ce-nionochlorobiphenyl
13C12-tetrachlorobiphenyl
13C!2-decachlorobiphenyl
Total spike
level (pg)
0.52
0.50
0.52
0.52
0.76
0.87
0.96
1.30
2.30
2.50
2.10
2.60
5.30
10.20
Spike 1 (%
Uncorrected
60.2
69.0
73.5
75.0
79.7
85.8
84.0
81.2
89.2
88.2
69.9
77.8
9.1
12
72.9
89.9
78.9
80.6
8.6
11
recovery)
Corrected"
82.6
94.6
100.8
83.5
88.7
95.4
93.4
90.3
113.0
111.8
88.6
94.8
10.2
11
_
-
-
-
-
-
Spike 2 (%
Uncorrected
67.7
73.6
77.5
77.6
80.7
85.0
85.0
81.3
89.3
87.6
71.9
79.7
6.8
9
75.1
87.0
76.0
79.4
6.6
8
recovery)
Corrected
90.1
98.0
103.3
89.2
92.9
97.7
97.7
93.5
117.5
115.3
94.6
99.1
9.5
10
_
-
-
-
-
-
Blank
NDb
ND
ND
ND
ND
ND
ND
ND
ND
ND
NDc
-
-
-
ND
ND
ND
-
-
-
a Corrected via surrogate response.
b Not detected.
c Not applicable.
-------
TABLE 25. RECOVERY
DATA FOR ALUMINA
CLEANUP
Total spike Spike 1 (%
Congener no.
1
3
7
30
50
97
143
183
202
207
209
X
Standard deviation
-p~ Relative standard
deviation (%)
211
212
214
X
Standard deviation
Relative standard
deviation (%)
PCB homolog level (pg) Uncorrected
Monochlorobiphenyl
Monochlorobipuenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
13C6-monochlorobiphenyl
13Ci2~tetrachlorobiphenyl
13Ci2~decachlorobiphenyl
0.52
0.50
0.52
0.52
0.76
0.87
0.96
1.30
2.30
2.50
2.10
2.60
5.30
10.20
63.1
60.0
67.9
NQC
67.2
70.4
69.4
75.8
76.8
77.3
74.0
70.2
5.9
8
64.8
69.1
83.2
72.4
9.6
13
recovery)
Corrected"
97.1
92.2
104.8
NQ
97.2
101.9
100.4
109.7
92.2
92.9
88.9
97.8
6.5
7
-
-
-
-
-
Spike 2 (%
Uncorrected
61.1
58.4
66.4
NQ
66.3
68.3
67.5
75.1
75.3
76.8
78.3
70.1
6.9
10
60.7
64.9
84.2
69.9
12.5
18
recovery)
Corrected
101.0
96.2
109.4
NQ
102.2
105.4
104.2
115.8
89.5
91.2
93.0
100.8
8.4
8
-
-
-
-
-
Blank
NDb
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
U
-
-
ND
ND
ND
-
-
-
a Corrected via surrogate response.
b Not detected.
c Large background
d Not applicable.
signal prevented quantitation
of this
compound.
-------
The preliminary data presented here contain an apparent anomaly: the
low recovery of the 13C-tetrachlorobiphenyl surrogate (Congener No. 212) from
the Florisil column cleanup. These two data points contribute substantially
to the imprecision of the surrogate recoveries and induce some very high (130
to 177%) corrected recoveries for the tri- through hepta- compounds. The ex-
periment should be repeated.
VALIDATION OF THE PRODUCT AND PRODUCT WASTE METHOD WITH INDUSTRIAL SAMPLES
Strategy
Selected samples, obtained from industrial sources, were subjected to a
variety of sample preparations as listed in Table 15 and then analyzed by
CGC/EIMS. This section presents the results of this preliminary validation
and, where possible, compares our values with those of previous analyses of
the same sample. The results for quality control samples are also reported.
The most extensively studied matrix was the CMA-A chlorinated benzene
waste stream sample. This particular sample was chosen because of the wide
distribution of PCB homologs (mono- through decachlorobiphenyls). Sample
preparation with this matrix included simple dilution, treatment with sul-
furic acid, Florisil, and saponification with ethanolic potassium hydroxide.
The CMA-A samples were analyzed in duplicate in two sets of experiments. The
11 PCB congeners used for calibration purposes were spiked into the CMA-A
matrix for standard addition experiments. Blind spiked samples and quantita-
tion standards, prepared by the MRI quality control personnel as analytical
performance checks, were analyzed along with the other samples.
First Sample Set
Tables 26 and 27 present the uncorrected and corrected concentrations
found for CMA-A samples in preliminary studies of the application of the pro-
posed methods for commercial products and product wastes. Sample 10 was
analyzed without surrogates to approximate the analytical procedure used by
most other laboratories. As anticipated, the uncorrected values compare well
with 20A and 20B, while the corrected values are slightly lower than the
values for 10. Both corrected and uncorrected values for the duplicate sam-
ples 20A and 20B are in agreement. The values for samples 10, 20A, and 20B
average about 400 (Jg/g. These values are higher than the mean of 280 |Jg/g
reported in the CMA round robin but are in good agreement with the values
(402 |Jg/g) reported by the sample supplier (Appendix E of Pittaway and Horner,
1982). The homolog distribution of our data agrees in general with the
CMA data and the data that accompanied the samples.
Sample 110 (CMA-E) was determined to contain about 18 pg/g PCB (Table
28) mostly as the dichloro homolog. These results are slightly higher than
the CMA round robin data, which had a mean reported value of 9 pg/g. The
isomer distribution agrees with most of the CMA round robin data (Pittaway
and Horner, 1982).
45
-------
TABLE 26. UNCORRECTED PCS CONCENTRATIONS (pg/g) IN CMA-A SAMPLES
Congener
no.
1
3
7
30
50
97
143
183
202
207
209
Total
211
212
214
PCB
homolog
1
1
2
3
4
5
6
7
8
9
10
1
4
10
10
Dilution,
no surrog.
9
19
64
55
60
50
56
60
0
0.9
9.3
414
NSa
NS
NS
20A
Dilution
11
21
70
52
63
40
48
84
0
0
20
408
96b
108
154
20B
Dilution
10
19
64
49
55
36
38
68
0
0
20
358
94
97
152
a No surrogates added.
b Surrogate recovery (percent).
46
-------
TABLE 27. CORRECTED PCS CONCENTRATIONS (Mg/g) IN CMA-A SAMPLES
Congener
no.
1
3
7
30
50
97
143
183
202
207
209
Total
PCB
homo log
1
1
2
3
4
5
6
7
8
9
10
10
Dilution,
no surrog.
NSa
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
20A
Dilution
11
22
73
49
58
37
44
78
0
0
13
385
20B
Dilution
11
21
68
50
57
37
39
70
0
0
13
366
a No surrogates added.
47
-------
Second Sample Set
CMA Product Waste Samples--
The corrected and uncorrected concentrations of the PCB homologs for
duplicate CMA-A samples from a more extensive study are presented in Tables
29 and 30. Sample 2005 was spiked only with the internal standard so that
any interferences corresponding to the 13C-labeled PCBs could be measured.
Samples 2010 and 2020 are duplicate samples of CMA-A. The four surrogate
compounds were added to approximately 0.1 g of each sample. The mixture was
diluted to 1.0 ml and the internal standard added. Sample 2025Q is a sample
that was submitted for PCB analysis by the MRI quality control department.
This sample was weighed by QC personnel and the final preparation completed
as described for the previous samples. The MRI QC coordinator calculated the
final concentration for 2025Q from the extract concentration of each PCB
homolog and weight of the CMA-A sample recorded in the QC laboratory record
book. The surrogate-corrected values reported for samples 2010 through 2025Q
are in good agreement with the total PCB concentration and homolog distribu-
tion reported in the CMA round robin (Pittaway and Horner, 1982).
Tables 31 and 32 present the data from a standard addition experiment
with the CMA-A sample matrix. The 11 PCB congener calibration standard was
added to three separate aliquots of the CMA-A matrix to give spike levels
ranging from approximately 20 to 100 |jg of the monochlorobiphenyl and 50 to
200 (Jg of decachlorobiphenyl. Samples 2030, 2040, and 2050 were prepared in
the analytical laboratory. Sample 2060Q was prepared as a blind spike of the
CMA-A matrix by MRI quality control personnel. The uncorrected amount found
did not increase linearly with the spike level. In fact, at the highest spike
level (Sample 2050) the amounts found for each homolog were less than the
spike. No explanation is immediately available for this data trend, although
the low recoveries of the 13C-octa- and tetrachlorobiphenyl surrogates indi-
cated that the data are at best marginally valid.
Tables 33 and 34 present data for CMA-A samples that were subjected to
three different cleanup methods (concentrated H2S04, Florisil column chro-
matography, and saponification with alcoholic KOH). The data from the sul-
furic acid cleanup was difficult to interpret because of interferences. As
noted previously (Erickson and Stanley, 1982), the acid cleanup results in
large losses of lower chlorinated PCB homologs. The poor recoveries of the
surrogates shown in Table 33 are clearly outside of the QC criteria in Sec-
tion 14.2.2 of Appendix B and indicate that the analyses are invalid. These
results would not be reported as analyses for compliance with the proposed
regulation.
All of the blank samples (2001, 1080, 2100, and 2120) were analyzed
along with the sample discussed above and found to contain no detectable
PCBs.
48
-------
TABLE 28. UNCORRECTED AND CORRECTED PCB CONCENTRATIONS (pg/g)
IN CMA-E SAMPLE (DILUTION PREPARATION)
Congener
no.
PCB
homolog
110
Uncorrected
110
Corrected
1
3
7
30
50
97
143
183
202
207
209
Total
211
212
214
1
1
2
3
4
5
6
7
8
9
10
1
4
10
1.2
1.8
10.5
0
0
0
0.02
0
0.05
0
0.06
13.4
103/91
151
1.5
2.4
13.8
0
0
0
0.02
0
0.03
0
0.04
17.7
b
a Surrogate recovery (percent).
b Not applicable.
c Samples run twice on magnetic sector instrument for low and high masses.
Congener no. 212 monitored in both runs.
49
-------
TABLE 29. UNCORRECTED PCB CONCENTRATION ((Jg/g) IN THE CMA-A
SAMPLE MATRIX (INTERNAL STANDARD CALCULATION)
PCB
homolog
CGC/EIMS analysis date
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
Total PCB
Recovery
13C6-monochlorobiphenyl
13C12-tetrachlorobiphenylc
13C12-octachlorobiphenyl
13C12~decachlorobiphenyl
CMA-A CMA-A CMA-A
2005 2010 2020
8/4/82 8/4/82 8/5/82
26
35
17
20
32
29
18
5.4
2.6
12
197
(%) of Surrogate
NSa
NS
NS
NS
23
28
14
31
29
23
12
4.1
2.2
10
176
Compounds
64
96
73
68
37
41
46
33
29
21
12
3.4
2.0
9.7
234
84
96
67
69
CMA-A
2025
8/5/82
40
48
50
36
31
22
14
4.2
3.5
11
260
89
101
72
73
a NS = no surrogate added.
b Final concentration determined from sample weight recorded by QC
coordinator.
c 302 Daltons used for quantitation.
50
-------
J8KCJOHS g'ffi'CAL L!^
801 L ST., S.W., TS-7^.;
KWSKINGTON, D.C. 2Q«Q
TABLE 30. CORRECTED PCB CONCENTRATION ((Jg/g) IN THE
CMA-A SAMPLE MATRIX
PCB
homolog
CGC/EIMS analysis date
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
Total PCB
CMA-A
2010
8/4/82
37
44
15
33
30
24
16
5.4
3.1
15
223
CMA-A
2020
8/5/82
44
48
47
34
30
21
18
4.9
3.0
14
264
CMA-A,
2025QD
8/5/82
44
53
49
34
31
22
19
5.7
4.8
16
280
a NS = no surrogates added.
b Final concentration determined from sample weight
recorded by QC coordinator.
51
-------
TABLE 31. UNCORRECTED PCB CONCENTRATION (pg/g) OF SPIKED CHA-A SAMPLES DETERMINED RY THE
INTERNAL STANDARD QUANTITATION METHOD
Ln
,
CMA-A 2030
Total sample
PCB homolog concentration
CGC/EIMS analysis date
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
Total PCB
13Cg-monochlorobiphenyl
13Ci2-tetrachlorobiphenyl
13C 12 -Octachlorobiphenyl
13Ci2-decachlorobiphenyl
8/5/82
60
56
65
47
48
40
40
46
51
60
513
89
94
62
65
CMA-A 2040
Spike
level
Total sample
concentration
CMA-A 2050
Blind quantitat ion
CMA-A 2060Q standard
Spike Total sample Spike Total sample Spike Total samp
level concentration level concentration level eonceiitr_at
8/5/82
20
10
10
15
17
19
25
45
49
42
252
80
58
75
55
58
48
58
82
93
110
717
Recovery
79
93
56
57
49
25
25
36
42
46
62
110
120
100
615
(%) of
8/6/82
92 100
58 51
39 51
43 75
64 86
61 95
87 130
100 230
130 250
140 210
814 1,280
surrogate compounds
76
84
41
48
8/6/82 fi
100 82 140
69 42 53
44 42 87
50 61 110
73 70 140
67 77 160
87 100 340
110 180 560
140 200 530
140 170 430
920 1,020 2,550
93
93
53
64
ile Spike
ion level
i/6/82
184
94
94
137
157
173
234
414
450
369
2,306
88
88
78
79
a Concentration in ng/ml rather than pg/g since this sample was prepared by dilution of stock solutions of standards by QC personnel.
b 302 Dal.tons used for quantisation.
-------
TABLE 32. CORRECTED PCB CONCENTRATION (pg/g) OF SPIKED CHA-A SAtlPLES DETERMINED BY SURROGATE RECOVERY CORRECTION
PCB homolog
CGC/EIHS analysis date
Honochlorobiphenyl
Dichlorobiphenyl
Trichl.orobiphenyl
Tetrachlorobiphenyl
Penta ch lorobipheriy 1
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
Total PCB
CHA-A 2030
Total sample
concentration
8/5/82
67
63
70
50
51
43
64
74
81
91
650
Spike
cone.
20
10
10
15
17
19
25
45
49
42
250
CHA-A 2040
Total sample
concentration
8/5/82
100
74
80
58
63
52
100
150
170
180
1,030
Spike
cone.
49
25
25
36
42
46
62
110
120
100
620
CHA-A 2050
Total sample
concentration
8/6/82
120
76
46
52
77
72
210
250
330
280
1,510
Spike
cone.
100
51
51
75
86
95
130
230
250
210
1,280
CHA-A 2060
Total sample
eoncentration
8/6/82
110
74
47
53
78
72
160
210
270
220
1,290
Blind quantitation
standard 2070Q
Spike
cone.
82
42
42
61
70
77
100
180
200
170
1,020
Total sample
concentration
8/6/82
160
60
99
130
. 160
190
430
720
680
540
3,190
Spike
cone.
184
94
94
137
157
173
234
414
450
369
2,310
Concentration in ng/ml rather than pg/g since this sample was a blind quantitation sample
-------
TABLE 33. PCB CONCENTRATION (pg/g) OF CMA-A SAMPLES TREATED WITH DIFFERENT
CLEANUP PROCEDURES (INTERNAL STANDARD QUANTITATION)
CMA-A 2090 CMA-A 2110 CMA-A 2130
PCB homolog acid cleanup Florisil cleanup alcoholic KOH cleanup
CGC/EIMS analysis date
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorpbiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
Total PCB
13C6-monochlorobiphenyl
13Ci2-tetrachlorobiphenyl
13Ci2~octachlorobiphenyl
13C12~decachlorobiphenyl
8/9/82
NDa
4.4
0.4
25
19
7.9
5.9
2.4
38
16
119
Recovery
74
0
115
173
8/9/82
4.4
14
31
18
17
5.6
2.2
6.0
2.4
9.5
110
(%) of surrogate compounds
8
0
97
129
8/9/82
31
44
44
25
20
6.3
3.8
2.6
2.6
6.4
186
145
367
110
64
a ND = not detected.
b 302 Daltons used for quantitation.
54
-------
TABLE 34. PCB CONCENTRATION (M8/g) OF CMA-A SAMPLES TREATED WITH
VARIOUS CLEANUP PROCEDURES (SURROGATE COMPOUND CORRECTED)
CMA-A 2090 CMA-A 2110 CMA-A 2130
PCB homolog acid cleanup Florisil cleanup alcoholic KOH cleanup
CGC/EIMS analysis date
Monochlorobiphenyl
Dichlorobiphenyl
Trichlorobiphenyl
Tetrachlorobiphenyl
Pentachlorobiphenyl
Hexachlorobiphenyl
Heptachlorobiphenyl
Octachlorobiphenyl
Nonachlorobiphenyl
Decachlorobiphenyl
Total PCB
8/9/82
NDa
30
0.3 (0.2)b
17 (11)
13 (8.3)
5.3 (3.5)
2.6
1.1
17
3.9
90 (78)
8/9/82
28
86
200 (16)
110 (9.3)
110 (8.9)
3.5 (2.2)
1.2
3.1
1.2
3.1
546 (159)
8/9/82
11
15
15 (20)
8.4 (11)
6.8 (9.0)
2.9 (2.9)
1.8
1.2
1.2
4.2
68 (77)
a ND = not detected.
b 13C!2-tetrachlorobiphenyl was not quantifiable due to interferences. The values
reported were calculated using ^Cg-monochlorobiphenyl. Values in parenthesis
were calculated using 13C12-octachlorobiphenyl.
55
-------
DCMA Pigment Samples--
Eight DCMA pigment samples were analyzed following the preparation de-
scribed in the experimental section (Table 15). The results are presented in
Table 35. The diarylide yellow pigment (DCMA-1) was analyzed in duplicate
and as a blind spike supplied by the MRI quality control department. This
sample is reported to contain 3,3'-dichlorobiphenyl at levels of approximately
70 |Jg/g (Dry Colors Manufacturing Association, 1981). No analyte or surrogate
PCBs were detected in the duplicate 1-g samples of the pigment and a known
spike of the sample. The lack of detected PCBs indicates a loss of analytes
in the sample preparation. The CGC/EIMS analysis of a sample of the yellow
pigment spiked by MRI quality control personnel yielded an uncorrected con-
centration of 76 M8/8 °f 3,3'-dichlorobiphenyl based on the internal standard
quantitation and a corrected concentration of 63 M8/8> based on 12070 recovery
of the 13C6~4-monochlorobiphenyl surrogate. The level of the 3,3'-dichloro-
biphenyl added by the QC personnel was reported to be 60 P8/8- Hence, the
total dichlorobiphenyl concentration should have been approximatey 130 |Jg/g
(70 ng/g endogenous plus 60 (Jg/g added).
The phthalocyanine green pigment (DCMA-4) was also analyzed in duplicate
following dissolution and fractionation with a Florisil column. This pigment
reportedly contains only decachlorobiphenyl at approximately 40 (Jg/g based on
the results of the DCMA round robin study (Dry Color Manufacturing Associa-
tion, 1981). Our analysis of duplicate samples yielded uncorrected concen-
tration levels of 24 and 27 (Jg/g of decachlorobiphenyl by the internal quanti-
tation method. The corrected concentration for both samples was 13 (Jg/g with
recovery of the 13Ce-decachlorobiphenyl surrogate at 190 and 210%.
Phthalocyanine blue (DCMA-8) was also analyzed as a single sample.
Pentachloro- and hexachlorobiphenyls were detected but the concentrations
were below the quantitation limits for that particular day. The total PCB
concentration of this pigment, as discussed in the results of the DCMA round
robin (1981), is reported to be 90 |Jg/g.
The DCMA pigment sample analyses did not produce valid results. These
data suggest that further development or validation of extraction/cleanup pro-
cedure would be necessary to provide acceptable PCB analyses of these samples.
All of the blank samples (2001, 2080, and 2100) analyzed along with the DCMA
samples were found to contain no detectable PCBs.
DISCUSSION
The determination of PCBs is a complex problem. The inaccessability of
standards for all 209 congeners has traditionally been circumvented by the
use of commercial mixtures (e.g., Aroclors) as standards. Quantitation has
often been addressed in terms of relating the analyte to an Aroclor standard
to give a "total PCB" concentration. Determination of PCBs synthesized as
by-products in commercial products or product waste presents three special
problems: (a) the analyte does not generally resemble a commercial PCB mix-
ture, so quantitation against Aroclor standards would be incorrect; (b) the
matrix often contains high concentrations of other chlorinated organics which
are not easily separated during a cleanup procedure and which interfere with
the qualitative and quantitative analysis; and (c) the matrix is undefined
and can include gases, liquids, or solids of any purity and complexity.
56
-------
TABLE 35. RECOVERY (%) OF CARBON-13 LABELED SURROGATE COMPOUNDS FROM DIARYLIDE YELLOW
AND PHTHALOCYANINE BLUE AND GREEN PIGMENTS
PCB
surrogate
13C6-Monochlorobiphenyl
13C12-Tetrachlorobiphenyl
13C12-Octachlorobiphenyl
13C12-Decachlorobiphenyl
DCMA-1
21403
ND8
ND
ND
ND
DCMA-1
21503
ND
ND
ND
ND
DCMArl
216?
ND
ND
ND
ND
DCMA-1
2170QC
120
ND
200
250
DCMA^A
2175d
ND
ND
120
190
DCMA-:4
2180
ND
ND
107
210
DCMA-8
21906
ND
94
92
150
DCMA-8
2200Q
12
52
71
77
a Samples 2140 and 2150 are duplicates prepared by the DCMA-B method.
b Sample 2160 was spiked with 50 pg/g of 3,3'-dichlorobiphenyl and prepared by the DCMA-B method.
c Sample 2170Q was spiked by MRI quality control personnel with 3,3'-dichlorobiphenyl and was prepared
by the DCMA-B method.
d Samples 2175 and 2180 are duplicates prepared by the DCMA-B method.
e Sample 2180 was prepared by the DCMA-A method.
f Sample 2200Q was weighed by MRI quality control personnel. An unknown mass of sample was supplied for
preparation by the DCMA-A method.
g The four surrogate compounds were added but not detected.
-------
In this situation, analytical methods require a different philosophy
than the classic approach for a single analyte in a defined matrix where all
steps, reagents, and apparatus are specified. The method proposed here leaves
many of the analytical steps to the discretion of the analyst while ensuring
the reliability of the results with a strong quality control program. Thus,
an analyst familiar with general analytical techniques for a product, may read-
ily adapt in-house extraction/cleanup procedures to incidental PCB analysis.
Even when the recoveries are not optimized, the l3C-labeled surrogate recov-
eries will mimic those of the analyte PCBs. As long as the 13C recovery sur-
rogates are thoroughly incorporated, their recoveries can be used to derive
corrected analyte PCB concentrations.
Several of the method validation analyses presented above, especially
Tables 33 and 35, illustrate the importance of the recovery surrogates in QC.
The techniques employed are common methods validated for PCB analysis by other
laboratories without the 13C-surrogate data. Analyses of this type may have
been used by a testing laboratory and erroneous results reported.
The complexity of the matrix and the high probability of chlorinated or-
ganic interferents precluded the use of GC/ECD. The best available technique
is GC/EIMS. During the validation work presented above, the anticipated dif-
ficulty of qualitatitve and quantitative data interpretation was confirmed.
In addition to the inherent problems resulting from extrapolation from a stan-
dard to several analytes, interpretation of the complex peak clusters is a
tedious, subjective, and error-prone process. The volume of data for one
sample is staggering; for sample 2110, 286 peaks were identified and inte-
grated in the PCB mass chromatograms as shown in Figures 6 through 16. Of
these, 58 peaks met the qualitative criteria and were identified as PCBs.
Clearly different analysts will obtain different results for those peaks
which marginally fit the qualitative criteria. This very high data density
relative to other common GC/MS analyses has a much higher potential for error
and mistakes. In addition it should be noted that, for many of the samples
analyzed in this study, the data interpretation is more time-consuming than
the rest of the analytical process.
The integration methods are also prone to error. Integration is always
conducted interactively with the mass spectrometric data system, either man-
ually or automatically. The selection of baseline criteria, background sensi-
tivity, integration method (valley-to-valley, baseline-to-baseline, etc.),
and retention window all affect automatic quantitation. The position of the
cursor and integration method affect the manual quantitation results.
The day-to-day instrumental variability with quadrupole systems also ap-
pears to adversely affect data quality, despite tight calibration specifica-
tions. The magnitude of this error soruce should be further documented.
The above discussion presents some of our understanding of some of the
major problems with analysis for by-product PCBs. Further work will be de-
voted to characterizing and reducing these problem areas. Even with forsee-
able improvements in the method, the data for by-product PCBs in many com-
mercial product and product waste samples will exhibit low precision and
accuracy.
58
-------
VO
BIG DATA: 4901IIO$J5 01
08/09/82 16:20:00 CALI: MIDCAIJWOVI fl4
SAIIPLE: SAMPLE 82110 QIA-A FLORISIL 1/IOOIL IUIIIU
RANGE: G 1.1750 LABEL: II 0. 4.0 GUAM: A 9. 1.0 RASE: U 2«. 3
993
SCAIIS 1 in 1750
200
3:19
6:20
800
12:40
1880
I5:'JO
1200
19:08.
1409
22:10
1660
25:2«
SCA1I
TIHF.
Figure 6. Reconstructed ion chroraatogram for SIM analysis of the CFA-A sample No. 2110.
-------
37.9-1
188
29. H
MASS CIIKOIIATOGRAIIS DATA; 190III09V5 Bl
08/09/82 16:20:00 CAM: IIIDTALIMWI fl1
SAMPLE: SAMPLE S2M0 CIIA-A FIOR1S1L I/IOIUL IULHU
RANGE: G 1.1750 LADEI.: H 9, 4.0 I3UAH: A 9. 1.0 BASE: U 29. 3
832
SCAHS 709 TO 900
916170.
im
11:05
12:40
850
13:27
900 sati
14:15 THIJ-:
Figure 7. SIM ion plots for monochlorobiphenyls (188 and 190 Daltons) and the C^-
monochlorobiphenyl surrogate (194 Daltons) in CMA-A sample No. 2110.
-------
100.8-!
222.
45.3-1
224 .
MASS CIIBOIIATOGRA11S DATA: 1901II99V5 Bl
88/09/82 16:20:00 CALI: IIIDCAUHW'Jl Bl
SAMPLE: SAIfPLE 82110 dlA-A FLORISIL t/IODfL 1UI.IHJ
RAIIGE: G 1.1759 LABEL: II 8. 4.0 QtJAII: A 0. 1.9 BASE: U 20. 3
SCAMS 700 TO 1100
1080
15:59
1099
154001^1.
222.RP6
* 0.5C3
6983P8.
224.607
* 9.SC9
1059
16:37
1109 SCAN
17:25 Tllll-
Figure 8. SIM ion plots for dichlorobiphenyls (222 and 224 Daltons) in CMA-A sample
No. 2110.
-------
NJ
MASS CHBOHATOGRAHS DATA: 490IH09V5 til
08/89/82 16:20:80 CAL1: tilDTALII09VM M
SAIIFLE: SAIIPI.E A2II0 QIA-A FLORISIL 1/I0DIL IULIIU
RAIIGE: G I.I750 LABEL: \\ 9. 4.9 01)All: A 0. 1.0 BASE: U 29. 3
SCAIIS 850 TO
180.0
256 .
89.6-
258
1093
972. 99?
1017
1093
1PI6
850
13:27
900
14:15
950
15:92
=»f—f r—-r- i 1
1000 1050
15:50 16:37
1100
17:25
1156120.
256.077
* 0.500
I304570.
258.077
* 0.599
1150 SCAH
18:12
Figure 9. SIM ion plots for trichlorobiphenyls (256 and 258 Daltons) in CMA-A sample No. 2110.
-------
71.3n
299 _
MASS QIBOIIMOCRMIS DATA: 499III99V5 81
98/99/82 16:20:90 CA1.|. HIBCALIW9tfl 01
SAMPLE: SAMPLE 82119 CIIA-A FLORISIL 1/I0DIL 1ULUU
BAHGE: G 1.1759 LABEL: II 9. 4.9 GUAII: A 9. 1.9 BASE: U 29. 3
1222
SCAHS 1959 TO 1350
109.9-1
292 .
39.7-1
298 _
62.7-1
394 .
1659
16:37
1158
18:12
1269
19:99
1259
19:47
1399
29:35
217296.
:>39.W»7
.'- 0.509
316621.
292.987
0.590
137728.
298.989
9.599
217344.
304.991
9.599
1350 SCAM
21:22 TIME
Figure 10. SIM ion plots for tetrachlorobiphenyls (290 and 292 Daltons), 3,3',A,A'-tetrachloro-
biphenyl-dg (298 Daltons), and the 13C;L2-tetrachlorobiphenyl surrogate (304 Daltons) in CMA-A
sample No. 2110.
-------
109.0-1
326 .
81.3-1
328
MASS CIIIiOIIATOGRAIIS DATA: 190III09U5 HI669
08/89/82 16:28:80 CALI: IIIDCALII99VI 11
SAMPLE: SAMPLE 82II8 CIIA A FLORISIL l/IODIL 1UIIIU
RAHGE: C 1,1758 LABEL: II 8. 4.8 OIJAII: A 8. 1.8 BASE: U 20, 3
12 7
SCAIIS !208 TO 15QO
1200
19:89
15897R.
326.897
* 8.588
129288.
328.898
* 8.588
1258
19:17
1300
20:35
14Q8
22:18
1158
22:57
1563 SCAH
23:15 TIIIE
Figure 11. SIM ion plots for pentachlorobiphenyls (326 and 328 Daltons) in CMA-A sample
No. 2110.
-------
97.4-1
»ASS aiROIIATOCRAm DATA: 49eiII0flV5 81
68/09/82 16:29:89 CALI: IIIDCALII03!M B1
SAIIFLE: SAIIPLE H2II9 CIIA-A FLOItlSIL l/IODIL IULUU
RANGE: 6 1.I759 LADEL: II 9. 4.9 OUAII: A 9. 1.9 CASE: U 29. 3
SCANS 1259 TO 1509
369 .
180.0-1
362 .
1259
19:47
1359
21:22
1499
22:19
1459
22:57
150784.
309.188
* 9.599
154889.
302.198
9.599
1593 SCAM
23:15 TIME
Figure 12. SIM ion plots of hexachlorobiphenyls (360 and 362 Daltons) in CFA-A sample No. 2110,
-------
OX
80.8-1
394
J.O-,
IttSS OIR011ATOGRAIIS DMA: 4901H00V5 81 S€MIS 1358 TO I5TO
08/09/82 16:20=00 CAll: IIIDCALIWWI D1
SAIIFLE: S.MIPLE I2H0 CHA-A FLORISIL 1/I0DIL lULIIU
RAHGE: G 1,1750 LABEL: II 0. 4.0 (IIJ.AII: A 0. I.0 ItASE: U 20. 3
1379
1516
45588.
391.118
0.508
56384.
396.118
0.590
1359
21:22
1559 SCAH
2-1-.32 THIE
Figure 13. SIM ion plots of heptachlorobiphenyls (394 and 396 Daltons) in CMA-A sample No. 2110.
-------
70.3-1
428 _
IIASS aiBOIIATOGRAIIS DATAt 490IHB9V5 Of
08/B3/82 16:20:09 CALIs HIDCAUieOVI HI
SAJIPLEt SA11FLE 82110 CIIA-A FLOP.ISIL I/1601L 1ULUU
RAIIGE: G 1.1750 LABEL: II 0. 4.0 CUAII: A 0. 1.0 BASE: 0 20. 3
1532
SCAHS 1440 TO 1650
98.9-1
430 _
1W.0-
442
29728.
4:>8. 128
11856.
438.129
>. 0.500
42304.
163<, 442.132
"*•- * 0.509
1450
22:57
1509
23:45
1550
24:32
1609
25:20
1C50 SCAN
20:07 TIME
Figure 14. SIN ion plots of octachlorobiphenyls (428 and 430 Daltons) and the C^2~octa"
.chlorobiphenyl surrogate (442 Daltons) in CMA-A sample No. 2llO.
-------
OS
00
MASS dlKOIIATOCRAIIS DATA: 4901II09V5 01
08/09/82 16:20:60 CAI.I: MIDCALIIOWI HI
SAMPLE: SAMPLE 82110 CMA-A FLORISJL 1/UMHL 1ULIIU
RAIIGE: C 1,1759 LABEL: II 0. 4.0 QUAD: A 0. 1.0 HASH: U 20. 3
SCAII3 1559 TO 1650
89.5-,
461
1631
24256.
164.139
0.500
100.0-,
1576
466
27104.
466.139
* 0.500
1560
24t42
1580
25:01
1680
25:20
1620
25:39
1610
25:58
SCAN
TIME
Figure 15. SIM ion plots of nonachlorobiphenyl (464 and 466 Daltons) in CMA-A sample No. 2110.
-------
89.8-1
498
189.8-1
509 .
49.7-,
518.
I1ASS C1IB01UTOGRAIIS DATA: 490III09V5 BI669
88/09/82 16:26:09 CALI: IIIDCALIIOTVI til
SAIIFLEt SAMPLE 121 IB OIA-A FLOBISIL I/IODIL IULIHJ
BAHGE: G 1.1758 LABEL: H 8. 4.0 QUAth A 8. 1.8 BASE: U 29. 3
1669
SCAIIS 1659 TO I7G3
1650
26:87
1668
26:17
79232.
08.119
0.509
1696
1669
tgg^^L.
88I92.
499.589
* 0.589
1684
1669
1696
-/^
43840.
5^9.508
* 0.588
IC99
1678
26:26
16C0
26:36
1698
26:45
1700 SCAH
2P:K> TIME
13,
Figure 16. SIM ion plots of decachlorobiphenyl (498 and 500 Daltons) and the C12~
decachlorobiphenyl (510 Daltons) in CltA-A sample No. 2110.'
-------
SECTION 5
REFERENCES
Dry Color Manufacturers Association. 1981. An analytical procedure for the
determination of polychlorinated biphenyls in dry phthalocyanine blue,
phthalocyanine green, and diarylide yellow pigments. 1117 North 19th Street,
Arlington, VA 22209.
Erickson MD, Stanley JS. 1982. Midwest Research Institute. Methods of
analysis for incidentally generated PCBs literature review and preliminary
recommendations. Draft interim report no. 1. Washington, DC: Office of
Toxic Substances, U.S. Environmental Protection Agency. Contract 68-01-5915.
Haile CL, Baladi E. 1977. Midwest Research Institute. Methods for deter-
mining the total polychlorinated biphenyl emissions from incineration and
capacitor and transformer filling plants. Washington, DC: U.S. Environ-
mental Protection Agency. Contract 68-02-1780. EPA 600/4-77-048.
Pittaway AR, Horner TW. 1982. Heiden, Pittaway Associates. Statistical
analysis of data from a round robin experiment on PCB samples. Washington,
DC: Chemical Manufacturers Association report.
Roth RW, Keys JR, Chien DHT, et al. 1982. Midwest Research Institute.
Methods of analysis for incidentally generated PCBs—synthesis of 13C-PCB
surrogates. Draft interim report no. 3. Office of Toxic Substances, U.S.
Environmental Protection Agency. Contract 68-01-5915.
Stanley JS, Erickson MD. 1982. Midwest Research Institute. Peer review and
authors' replies to 'methods of analysis for incidentally generated PCBs--
literature review and preliminary recommendations.' Draft interim report no.
2. Washington, DC: Office of Toxic Substances, U.S. Environmental Protection
Agency. Contract 68-01-5915.
USEPA. 1979a (December 3). U.S. Environmental Protection Agency. Base/
neutrals, acids, and pesticides—method 605. 44 FR 69540.
USEPA. 1979b (December 3). U.S. Environmental Protection Agency. Organo-
chlorine pesticides and PCBs--method 608. 44 FR 69501.
USEPA. 1982 (June 8). U.S. Environmental Protection Agency. Polychlorinated
biphenyls (PCBs); manufacture, processing, distribution, and use in closed
and controlled waste manufacturing processes. FR 74 24976.
70
-------
APPENDIX A
SUPPLEMENTARY GC/EIMS DATA ON PCS CONGENERS
A-l
-------
The following data support the method validation section for gas
chromatography/electron impact mass spectrometry (GC/EIMS) of polychlorinated
biphenyls (PCB). Table A-l lists the average relative response factors (RRF)
for the 77 commercially available PCB congeners determined as four replicates.
Table A-2 presents results of the Student's t-test used to determine the sig-
nificance of differences for average RRFs for PCB homologs measured on a
single day versus multiple days. The data in Table A-2 indicate that only
the average RRFs for the heptachlorobiphenyl homolog are significantly dif-
ferent.
Table A-3 presents the results of the Student's t-test used to determine
the significance of differences for the average RRFs for the PCB homologs de-
termined with the quadrupole and magnetic sector mass spectrometers. All 77
PCB congeners were determined in a single day for each of the instrument stud-
ies. This comparison indicates that the average RRF values are significantly
different, which was expected. However, the relative standard deviations are
not significantly different, indicating that the selection of the calibration
standards is appropriate. These conclusions are discussed more fully in the
text.
Table A-4 presents results of the Student's t-test used to determine
significance of differences for the RRFs for the 11 congeners in Solution
No. 1, which was analyzed daily. An example of the data generated for multi-
ple analysis of Solution No. 1 is presented in Figures 1 to 23. This infor-
mation includes a capillary GC/EIMS chromatogram of Solution No. 1, the mass
spectra of each component in this solution, and a graphic illustration of the
distribution of several measurements of each congener about the average re-
sponse factor. It should be noted that the standard deviation and relative
standard deviation presented in these plots are different from that reported
in the text due to calculation of the standard deviation using N weighting
rather than the correct N-l weighting. All other standard deviations reported
in this document are based on the N-l weighting.
The relative retention times of the 77 PCB congeners with respect to
3,3',4,4'-tetrachlorobiphenyl-d6 determined with the Finnigan 4023 quadrupole
and the Varian MAT 311A mass spectrometers are presented in Table A-5. A
relative retention time unit of 0.01 (10 sec) is required for resolution of
two specific congeners based on the gas chromatography parameters used to gen-
erate these numbers.
A-2
-------
TABLE A-l. RELATIVE RESPONSE FACTORS FOR COMMERCIALLY
AVAILABLE PCS CONGENERS (QUADRUPOLE)
Congener
no.
1
2
3
4
5
7
8
9
10
11
12
14
15
18
21
24
26
28
29
30
31
33
40
44
47
49
50
52
53
54
61
65
66
69
70
72
75
77
Degree of
chlorination
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
Average relative
response factor
, 4.073
2.951
2.969
1.232
1.959
2.008
2.049
2.148
1.880
3.073
1.929
2.083
1.909
1.104
1.586
1.051
1.714
. 1.587
2.195
1.526
1.706
1.688
0.597
0.712
1.062
0.831
0.957
0.732
0.750
0.958
0.975
1.086
1.139
1.058
1.091
1 0.980
1.185
1.095
Standard
deviation
0.118
0.056
0.028
0.008
0.035
0.027
0.023
0.061
0.031
0.073
0.036
0.098
0.089
0.012
0.018
0.033
0.013
0.028
0.048
0.067
0.024
0.031
0.013
0.007
0.059
0.019
0.025
0.011
0.008
0.013
0.069
0.022
0.068
0.012
0.050
* 0.048
0.061
0.050
Coefficient of
variation (%)
2.905
1.894
0.956
0.646
1.803
1.366
1.134
2.846
1.658
2.363
1.877
4.702
4.686
1.089
1.110
3.105
0.731
1.733
2.188
4.418
1.409
1.863
2.152
0.946
5.591
2.245
2.574
1.504
1.006
1.344
7.094
1.994
5.966
1.110
4.548
4.870
5.113
4.595
(continued)
A-3
-------
TABLE A-l (continued)
Congener
no.
87
88
93
97
100
101
103
104
115
116
119
121
128
129
136
137
138
139
141
143
151
153
154
154
155
156
171
181
183
185
195
198
200
202
204
206
207
208
209
Degree of
chlorination
5
5
5
5
5
5
5
5
5
5
5
5
6
6
6
6
6
6
6
6
6
6
6
6
6
6
7
7
7
7
8
8
8
8
8
9
9
9
10
Average relative
response factor
0.617
0.611
0.574
0.719
0.727
0.653
0.566
0.824
0.853
0.785
0.762
0.948
0.499
0.431
0.689
0.533
0.433
0.462
0.419
0.490
0.473
0.549
0.221
0.511
0.587
0.599
0.346
0.383
0.380
0.336
0.263
0.262
0.301
0.250
0.221
0.193
0.237
0.259
0.213
Standard
deviation
0.011
0.005
0.010
0.008
0.003
0.004
0.009
0.025
0.061
0.013
0.022
0.020
0.005
0.004
0.016
0.008
0.008
0.026
0.010
0.005
0.013
0.050
0.001
0.010
0.011
0.044
0.002
0.009
0.010
0.006
0.003
0.008
0.007
0.007
0.007
0.003
0.008
0.003
0.006
Coefficient of
variation (%)
1.710
0.744
1.677
1.139
0.428
0.538
1.627
3.048
7.146
1.654
2.911
2.127
1.093
0.813
2.336
1.582
1.946
5.686
2.353
0.986
2.826
9.101
0.570
2.039
1.828
7.431
0.640
2.379
2.501
1.729
1.184
2.887
2.392
2.663
3.200
1.723
3.547
1.315
2.837
Relative to 3,3*,4,4'-tetrachlorobiphenyl-de- All relative response
factors were calculated as the average of four replicate measurements
made on the same day.
A-4
-------
TABLE A-2. STUDENT'S TWO-SIDED t-TEST TO DETERMINE SIGNIFICANT DIFFERENCES BETWEEN
QUADRUPOLE RESPONSE FACTORS CALCULATED ON THE SAME DAY VERSUS MULTIPLE DAYS
Ul
Number of
PCB homolog
Monochloro-
Dichloro-
Trichloro-
Tetrachloro-
Pentachloro-
Hexachloro-
Heptachloro-
Octachloro-
Nonachloro-
Decachloro-
isomers
3
10
9
16
12
13
4
6
3
1
Average RRF
from
replicate
measurements
3.331
2.027
1.573
0.950
0.720
0.513
0.361
0.253
0.229
0.213
Standard
deviation
0.643
0.447
0.341
0.175
0.120
0.078
0.024
0.030
0.034
0.006
Average RRF
from
single ,
measurement
2.739
2.048
1.592
0.946
0.725
0.500
0.308
- 0.224
0.188
0.179
Standard
deviation
0.254
0.322
0.289
0.189
0.127
0.096
0.025
0.039
0.030
Significant
t-Statistic at
1.478
-0.119
-0.131
0.0618
-0.1085
0.377
3.119
1.398
1.|J91
95% level?
No
No
No,
No
No
No
Yes
No
Na
a Four replicate measurements of the RRF were made for each isomer. For example, the three monochloro-
biphenyl isomers were measured four times each. Hence, the average RRF and standard deviation were
calculated from 12 distinct values.
b A single measurement for each of the 77 PCB congeners was completed in a single day. Hence, the
average RRF reported is the average of one measured RRF for each isomer within a homolog. For
example, the average RRF and standard deviation reported for the monochlorobiphenyl was calculated
from three distinct values.
c Single measurement.
d Cannot test significance of difference between single measurements.
-------
TABLE A-3. COMPARISON OF THE AVERAGE RELATIVE RESPONSE FACTORS (RRF) DETERMINED WITH QUADRUPOLE
(FINNIGAN 4023) AND MAGNETIC SECTOR (VARIAN MAT 311A) MASS SPECTROMETERS'"
Finnigan 4023
quadrupole MS
PCB homolog
Monochloro-
Dichloro-
Trichloro-
Tetrachloro-
Pentachloro-
Hexachloro-
Heptachloro-
Octachloro-
Nonachloro-
Decachloro-
Number of
isomers
3
10
9
16
12
13
4
6
3
1
RRF
2.739
2.038
1.592
0.946
0.725
0.500
0.308
0.224
0.188
0.179
Standard
deviation
0.250
0.32
0.29
0.19
0.13
0.10
0.025
0.04
O.g3
Varian MAT 311A
magnetic
sector MS
RRF
2.329
1.663
1.167
0.902
0.780
0.640
0.497
0.463
0.467
0.586
Standard
deviation
0.199
0.229
0.248
0.130
0.136
0.124
0.060
0.071
O.J05
RRFs significantly
different at the ,
95% confidence level
No
Yes
Yes
No
No
Yes
Yes
Yes
Yes
e
Variances significantly
different at the
95% confidence level
No
No
No
No
No
No
No
No
No
_e
a The RRF and standard deviation reported in this table for the quadrupole and magnetic sector mass spectrometers
were determined as single measurements of all congeners in a single day with each instrument.
b Student's two-sided t-test was used to determine significant differences of the RRFs.
c An F-test was used to determine significant differences of the standard deviations, where
F = (std dev!)2/(std dev2)2 with (n-1, n-1) degrees of freedom.
d Single measurement.
e Cannot test significance of difference between single measurements.
-------
TABLE A-4. STUDENT'S TWO-SIDED t-TEST TO DETERMINE SIGNIFICANT DIFFERENCES OF THE AVERAGE RELATIVE
RESPONSE FACTOR (RRF) FOR SOLUTION NO. 1 FOR REPLICATE ANALYSIS
ON A SINGLE DAY VERSUS SINGLE ANALYSES ON MULTIPLE DAYS
PCB
congener
no.
1
11
29
47
121
136 -
181
195
207
209
Replicate analyses
on single day
RRF
4.073
3.073
2.195
1.062
0.948
0.689
0.383
0.263
0.237
0.213
Standard
deviation
0.118
0.073
0.048
0.059
0.020
0.016
0.009
0.003
0.008
0.006
Single analyses,
on multiple days
RRF
3.241
2.538
1.899
1.015
0.959
0.683 -
0.374
0.275
0.269
0.230
Standard
deviation
0.201
0.161
0.100
0.059
0.043
0.058
0.035
0.028
0.032
0.027
t-Statistic
7.468
6.204
5.483
1.268
-0.479
0.186
0.662
-1.137
-2.479
-1.599
Significant differences
of RRF at
confidence
Yes
Yes
Yes
No
No
No,
No
No
Yes
No
95%
limit?
a The RRF and standard deviations were calculated from four replicate measurements completed in the same
day.
b The RRF and standard deviatons were calculated from seven single measurements from seven different days.
-------
TABLE A-5. RELATIVE RETENTION TIMES (RRT) OF 77 COMMERCIALLY AVAILABLE
PCB CONGENERS MEASURED VERSUS 3,3*4,4'-TETRACHLOROBIPHENYL-d6
DETERMINED WITH MAGNETIC SECTOR (VARIAN MAT 311A) AND
QUADRUPOLE (FINNIGAN 4023) MASS SPECTROMETERS
RRT
PCB congener no.
Monochloro-
1
2
3
Dichloro-
4
5
7
8
9
10
11
12
14
15
Trichloro-
18
21
24
26
28
29
30
31
33
Tetrachloro-
40
44
47
49
50
52
53
54
61
65
66
69
70
72
75
77
311A
0.403
0.481
0.474
0.518
0.598
0.559
0.590
0.563
0.521
0.649
0.660
0.616
0.677
0.665
0.762
0.685
0.729
0.745
0.719
0.641
0.741
0.760
0.870
0.838
0.814
0.811
0.746
0.804
0.763
0.720
0.898
0.822
0.905
0.800
0.880
0.853
0.816
1.002
4023
0.425
0.490
0.499
0.536
0.606
0.579
0.606
0.577
0.534
0.660
0.671
0.628
0.681
0.678
0.767
0.694
0.738
0.753
0.728
0.653
0.752
0.769
0.875
0.843
0.819
0.817
0.751
0.810
0.773
0.731
0.898
0.826
0.908
0.807
0.904
0.856
0.821
1.003
PCB congener no.
Pentachloro-
87
88
93
97
100
101
103
104
105
116
119
121
Hexachloro-
128
129
136
137
138
139
141
143
151
153
154
155
156
Heptachloro-
171
181
183
185
Octachloro-
194
195
198
200
202
204
RRT
311A
0.979
0.913
0.907
0.976
0.878
0.945
0.870
0.829
0.988
0.985
0.964
0.911
1.163
1.128
0.994
1.118
1.108
1.037
1.096
1.050
1.020
1.074
1.002
0.929
1.194
1.189
1.178
1.154
1.166
1.355
1.326
1.275
1.203
1.194
1.209
4023
0.978
0.915
0.908
0.979
0.884
0.945
0.874
0.836
0.987
0.986
0.965
0.914
1.156
1.127
0.996
1.115
1.103
1.038
1.093
1.051
1.021
1.073
1.004
0.931
1.188
1.187
1.174
1.148
1.161
1.351
1.317
1.265
1.199
1.188
1.203
(continued)
A-8
-------
TABLE A-5 (continued)
RRT
RRT
PCB congener no. 311A 4023 PCB congener no. 311A 4023
Nonachloro-
206
207
208
1.414
1.336
1.319
1.399
1.330
1.318
Decachloro-
209
1.453 1.440
A-9
-------
0:33:00
SAMPLE: PCB IHXTUKE—SOLUTIOII III
CO!!»S.: -1600 HIV
RAIitiE: C 1.1?CO
g»: CSS??!1765
IIJL [||J
TO
TO
I2fl0
I2CO
10-6 70EV .2IIA nD!i-HI( 10-2II-325-C/ "OII-COI"
LADEL: I! 0. 1.0 U.UAMs A 0. I.O J 0 BASE: U 20. 3
>
1
r-1
O
-HUM! 1.
^
c1 *-V96
(U
XI
p.
XI >, i-H
o a >>
H H
~* O
0.
i • i
59 c
M XI
0
CO
0)
1
^j
«
^
CM
*
CM
.-*.
0.
rH
C
(U
XI
p.
•H
(">
O
0
rH
XI
o
CO
4J
0)
PL.
?I2
rH
C
(LI
P.
•iH
O
H
O
rH
U
Cfl
s
1 1 I
^"I ^°
In vo"
•u -d~
H <^
CO
-* "
U
O
1
in"
**
_
CM
CM
. 1.251
\- —h
^ 599910
C
(ll
"p. 1.-1M
•H
Q
O
6, 6 '-Nonach]
""1
Zr
-------
59.6-
ii/n
G->/jy/:}2 IC: 12:Cfl * 1:5?
SAIirLE: PCB IHXIllllE—SOLUTION 111 lUL lilJ
OJiiDS.: -ITOO BIV IO-6 70EV .2IIA Ili'o-ljll IIO--2II-325-IO/ "Kil-CnL
I3BIAKCED (S ISO 2H Of)
DATA: 190in20yi>7 ()29«
I'M.I: CALF.20UI m
BA'JK H/K: I"')
lilC: 295 1.24.
•52
12G
99
173
IS8
189
9523*'.
129
. 110
ICO
100
Figure A-2. Electron impact ionization mass spectrum of 2-monochlorobiphenyl,
-------
100.0-1
59.0-
It'll
I;A-:S tt'KCTWjii
WJ/W/Kl Ui:12:00 * /:'jt
SAliri.!:: FOB IllXtlll!!:—-SDUMimi III HH. Itll
owns.: -imo tnv lo-c ;
ElillAliCHD IS 1511 2(1 01)
DATA: iyi)ll:2«Ui>7 B173
CAM-. CA1.K20V1 H3
99
100
ll l
Jlu-iiilu
120
110-211 325- It)/
C2
Jt3«
1DO
2«»
DAS!: II/'E: 222
UtC: 2GJG!}!).
?22
220
C329G.
Figure A-3. Electron impact ionization mass spectrum of 3,5-dichlorobiphenyl,
-------
50.0-
83
II.CS SPECTRI!il
KATA:
i:AM:
05/20/«2 10 :
-------
J00.0-1
50.0-
83
IIASS sr
05/20/82 10:12:00 + f):M
SAMPLE: PCD IIIXTIIISE—-5IJJ.IJTIOH Bl IUL II!I
CUiJDS.: -l'j'31) B1V 10-0 70liV ,21IA 1)115-IfM ltO-211-325 UV
B3WtlCn» (S I5B 211 OT»
DATA: 1WH:?OW7 IKS1
CALI: i:Al.i:2«»l 0.1
DASK II/E: ?:!*>
P.Ii;: 3MOJ2.
I 0
128
123
135
I?
4
I T
LjL
255
2?>2
39360
II/E U50 158 2««) 250 3S9
Figure A-5. Electron impact ionization mass spectrum of 2,2',4,4'-tetrachlorobiphenyl.
-------
100.On
59.8-
HAW s
8S/29/82 10:12:00.+ 1?.:RO
SAilPLC: rut! UlXTUr.E—-SOLUTIOil (II IWL IIU
conns.: -iwo nnv IO-G 7oi;v .2ii.A ofl'j-riii 110-211-325-io/ -OJI-
IS 1511 211 OT)
102
ro
III
189
171
163
D.A1.A: '1(J:MI.:20V"/ QTffi
IJAI.l: <:AI.I:2CVM in
2.90
n.Ast;
II 1C:
2f
lll^Ji|j.VL-r^r_|l)<
I ' I •' I"'
350
»/): \W> I'M 2W) 250 3«W
Figure A-6. Electron impact ionization mass spectrum of 3,3'A,4'-tetrachlorobiphenyl-d5.
-------
100.0-1
59.0-
99
J
HAS-; M'u
05/7:yi5:? 10:12:00 + 11:01
SAiiru;: res MiXTunii—-S«I.UTIOII BI sui. KM
CCSSUS.: -I5'J9 CIIV 10-0 7«I:V .2IIA l)U:i I'jll 110-211-323-10/ "IJJI--COI.
EIIIIACCI'D (S 15B 211 Of)
DATA: 1!>OIl:2c;Vlt/ (KIO'J
CAM: «JAI.F.2«IM 0.}
122
163
|R1
150
2H>
250
BASK \1,'f.: 3>
P.IO: IIW'
326
*" i
^Jlk-r^-
300
3078
Figure A-7. Electron impact ionization mass spectrum of 2,3',4,5',6-pentachlorobiphenyl.
-------
>
I—1
~J
50.0-
IIACS su;cit:)in
e!i/28/«2 10:12:00 > 12:86
irifi: rcu nixTunr—SOLUTION ni IUL nu
-I5b"0 CUV 10-5 70CV .2il\ 1)1)5-I'ill I
(S I5D 211 01)
9
It'Ji
128
2 0
2«0
DATA: 19H1E20U07 U7?0 PASF: ll/t: %«
CAM: CALI:20VI l)3 HIC: 323501.
10-211-325- ID/ "051-COI."
2
2LM
Iff?
* 1 • 1 l"^^~l • 1 — ^
250
31
?e
11
325
1 , .1 ,L.
t—'r— *i— | • |-*"l "i i | i | W-*
3tfO 350
9
1)
1
25661.
Figure A-8. Electron impact ionization mass spectrum of 2,2',3,3',6,6'-hexachlorobiphenyl.
-------
i
i-1
oo
100.0-
50.0
IIASS SPECTRUM
tfj/T.Vift IG-12-00 •>• I1-2I
SAurLE: run iiixiui::-—-souirio!! ni IIM. mi
COi!D:i.: -I'jW BIV 10-0 701-V .2IIA ll.Vi-IMI 110-211-325-IO/ "OSJ-COI."
HKIAKCED (S 15B 211 OT)
1C?
DAM: isi.»iE2f.v»7 o««i
i;AJ.l : CAI.I-20UI flj
BA';E ii/i-: :;y-3
ll^I I
UO.0-1
50.0-
LI
2CT
I2C
170
M
120
271
11
200
3»9
iW
32«
31D
97
220
249
3?'l
3CO
1 I ' ' ' ' '
300
400
ICOGt.
r iceo
Figure A-9.
Electron impact ionization mass spectrum of 2,2',3,4,4',5,6-heptachlorobiphenyl.
-------
I00.0-T
50.0-
95
I03
v^-J
2?3
50.0-
271
200
SI'ECTRWI
'«2 10:12:00 + |f>:H
[•: rc» iiixiuni-: — SUUITIOII in mi. IH.J
. : -1550 HIIV iO-O 70I-V .2IIA DilS-l'iSI II
CED (S I5II 2II OT)
IV
125 132 «?2
129 1
129 HO ICQ 1
3C
323
!f)0 3j| | |J t
380 320 310 3
PATA: l!HUI-2fitfl.«7 ll!)7l RAW: H/E: 170
•JA.I.I: CAI.F.2CM 03 ItlC: .'JJCB11.
0-2II-375-IO/ "Oil COI."
4^
2
I9I>
III OfVi
l| T_|.r M |'[ l|l|_.» [ l| l|.( , |
09 2C?0
B
R
,» -n In
229 219 269
393
1
-I, 1
I
Ll. . iL
60 300 1»50 120 4i9
16G7:
IGG;
Figure A-10. Electron impact ionization mass spectrum of 2,2',3,3',4,4',5,6-octachlorobiphenyl.
-------
100.0-1
50.0-
100.0-
50.0-
IIACS SiliCtilllil
05/20/1.7 lfi:42:OT ' 10-. fj
SAi!n.i;i rcii iiixiur.;;- — SULHTIUII ui tui. it.u
CMDS.: - TOO DIV IO-« 70RV .2IIA DU'i- riil 1 10-211-325- IO/ "051 -COI."
ElUIACCtD (S I5lt 2tl OTI
UG
DATA: l
Jl
ino
i |
190
2
luiiii L)
232
1
III f
W
Uit.. I
IG9 120 140 100
322
17661.
I 2?° 3?7
il.-..il-ii.TTf-i
320
3CO
30!)
4<
*?s-.
i-1
4Wi 120 -110 1CO t«S
Figure A-ll. Electron impact ionization mass spectrum of 2,2',3,3',4,4',5,6,6'-nonachlorobiphenyl.
-------
I)\U:
CAM: CAI.n2P.yi (13
liAG'J SPECTRUM
e5/?0/C2 10:12:00 + 17:13
SAiirti-: rcn iii.vnir.r- -soiunnii in nti. itu
OHMS.: -1558 HIV 10-6 70I;V .2IIA Dift-1511 110-211-325-IO/ "UJI-COL
(S I5H 211 OT)
211
IBO.O-
50.0-
II/E
198.0-
50.0-
170
1
96
L ll.
109
32
J
17
1?
1 f
1 ., J1J
'
., ?ft ,
I
1
I32
JL|lLj
•3
I
ISO
35(
i
I ,
0
ir,e
III
379
II,,
IM
L.
*
,l.l
f-
2
1 2?' 1
1 t , • .[II. lilij ,.1
2
1;
393
1
1
1
B
JU-,-.. , . ffi
0
,. f? ?!?.
2*6
II. 29? .
s~| j •
50 3f«
100
I.
|
I • 1
22016.
r 2281
ll/'fi
350
-150
ses
Figure A-12. Electron impact ionization mass spectrum of 2,2',3,3',4,4',5,5',6,6'-decachlorobiphenyl.
-------
mt:0i.ii HAW: 290 «nei;.coiir:0i.it IIASS: 2001
MUM) G TnTnAciii.ciiiHJmiBiYL. ISOKH: 0210
f:ti::D-ti n-iR.'.aii.o!:o--BiFi!i3M.. ISOIIKI: 11210
iAni;v/rn:.Ai:tA>/(AiiT./r.oi;.,\:iT.> (AV:
(i.OOO
% ST.DEY.-
O.OOD
i.o
I
N)
ho
l.Ct;0
O.S90
-x-—5:—Jc
IiATE
5/1-1/32
5/21/02
6/ 3/02
Fie re A-13 Response factor plotted on a day-to-day basis for the internal standard,
3,3' ,4,4'-tetrachlorobiphenyl-d6, in Solution No. 1.
-------
I
10
RBI'OilSli LIII:01.2i IIASS: 108 (i:i-:F.COIIP:QI. li
DIP : 2-IIOHI111I5.0KO-DI PIIRIVI.. J SCIfEP, II 1
P,tF:I)-G TtTPACIlLOnO-BirilB'.Y!.. ISOIIEB 112 U)
(AI!L'A./»',rf .AI:J-A)/(AIIT../nb'F.AIIT. ) (AV: '
5.GOO
S: 2:"«)
1.0CO
3.C80
2.G90
DATK 5/1-1/82
X
t
si.i;;;1.'.-
0.431
X ST.riEY.-
12.171
Sl'il -1.9
x ;
Figure A-14. Relative response'factor for 2-monochlorobiphenyl in Solution No. 1
calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis.
-------
I
ro
ni-:sroi!5F. 1.10:01,1. MASS: 222 im-:i:.o»ir:oi.ii iiAr,5: 2001
aiP:3.3>-nic!ii.nr.o-nin:i:.:ivi.. ISOIIHR nil
r.GI::D-C TimiACIIIJinO-BMIIBIYL. ISOIini! 0210
l!CT.AIIT.) (AV: 2.7;?2»
}
2.890
2.703
2.669
2.583
2.400
2.380
2.200
DATE 5/11/02
sr.ntv.-
0.206
Tt. ST.DEY.-
51II - l.t
5/21/02
€./ 3/32
Figure A-15. Relative response factor for 3,3'-dichlorobiphenyl in Solution No. 1
calculated versus 3,3',4,4'-tetrachlorobiphenyl-d^ on a day-to-day basis.
-------
RESPONSE i.in:0i.i. MASS: 25G
uip:2.i.5-TRiaii.oi:o-niniEiiYi.. isotiui! n29
REF:n-G TL:iR.\aii.ni;o-niriiiaiYi.. ISOIIEI: 11210
(ARP..A/l!rr.Ai:i:A)/(AIIT./m:r.A!IT.) (AY: 2.085I
2.2(in
2. «00
2.000
^>-
i •
Ul
1.999
1.883
1 7IM4
i
;
'
1
X
'
f
-
>
X
X
h
:
>
ST.D1-V.-
0. 163
? ST.IEY.-
ft. 138
Si«l =1.8
X
1
1
t
1
*
X
:
5/24/82
6/ 3/82
Figure A-16. Relative response factor for 2,4,5-trichlorobiphenyl in Solution No. 1
calculated versus 3,3',4,4'-tetrachlorobiphenyl-ds on a day-to-day basis.
-------
>
K>
Ri-srousr rin:0i.5i MASS: 202 (ni-:F.conr:oi.ii MASS: 200
aiP:2.2' .i.r-Tp.ir.Aaii.or.o-BiriiuiiYi.. ISOHHR BIT
nEF:D-0 rEJUAaiLPRO-niPIIFJJVI.. ISOWR 1J2IO •
(AR[-A/nB:.A!:i-AI/(Alir./!!l;l;.AIIT.) (AV: 1.0321
1.200
1.100
0.C90
DATK 5,'I'1/82
'•
5/2-1/82
ST.DI-V.-
0.050
A ST.DEY.-
5.5%
S!« -1.9
G/ 3/82
Figure A-17. Relative response factor for 2,2',4,4'-tetrachlorobiphenyl in Solution No. 1
calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis.
-------
nESPOIISK
I.IBrOI.C. MASS: 32G (l!F.F.COIir:01. !•
i.. isoni;n ni2i
..
(Ani-A/BF.P.Ai:nA»/(AIIT./riKF.AIIT.) (AV:
1.030
0.899
DATE 5/11/32
\
f
)
<
-_l _ — — -
>
(
-ST.IiHY.-
(1.052
2 ST. COY.-
>.477
St'il
5/2-1/82
G/ 3/82
Figure A-18. Relative response factor for 2,3',4,5',6-pentachlorobiphenyl in Solution No. 1
calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis.
-------
I
ro
co
1.10:01.?! IIASS: 3co u:i'F.co!ir:Gi.ii HASS:
air:2.2\3.3\6.6i-Mi:KAaiLcr.i)-BmiFjiYi.. ISOIIF.B« no
BF.F:0-G TE1KAClll.Oi:0-lliri!lilY!.. 1 SOI 11:1! (!2IO
(ARfiA/l!EF.Ar.!:A)/(AttT./IltH.AIIT.I (AV; 0.6C5I
O.PO'.I
o./oo
O.C09
0.580
DATE 5/14/82
5/2-1/82
ST. BEY.-
0.014
S ST.DEY.-
0.370
SHI - I.0
X
C/ 3/02
Figure A-19. Relative response factor for 2,2',3,',6,6'-hexachlorobiphenyl in Solution No. 1
calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis.
-------
ho
vo
n MD:OI.RI mss: 391 ii!nF.caiir:ni.ii MASS:
air:2.2'.3.-i./i'.5.G-iici>T/,nii.or.o-BiriiEiivi.. isonnn n mi
BEF:B-c TKTnAUii.ono-iiiriieivi.. isonia; 11210
URFA/RHF./.RCAVtAriT./r.tF.AIIT.) (AV: 0.377>
U.-1JW
0. 110
o.w -
0.300
0.380
0.370
0.360 -
8.350
0.310
8.330
o tin
X
,f
• i
• i
i
i
t
f
•
1
t
I
X
)
>
c
ST.M-V.-
< 0.027
S ST.IiEY.
7.035
5WI - 1.0
X
MX
X
X
5/2-1/82
G/ 3/82
Figure A-20. Relative response factor for 2,2',3,4,4',5,6-heptachlorobiphenyl in Solution No. 1
calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis.
-------
.1
OJ
o
r.Ksro;isn i.iD:oi.9i IIASS: 130 .
CIIP:2.2l.3.3'.1.'T.5.6-l)lJTAf:MLOno-l)II'lli;ilYI.. ISOIIRR II I95
REi::D--c TniiiAciiLor.o-Kiriii:>m. isounn 11210
(Ar.t:/./P,£F.AP.KA>/(AllT../P.\;i--.AllT.) (AY: 0.27fl)
0.30i>
o.:
0.269
0.250
o.;
sr.nr.v.-
0.021
z sr.ncv.-
7.072
1.0
X
DATE
5/14/82
5/2-1/82
C»/ 3/02
Figure A-21. Relative response factor for 2,2',3,3',4,4',5,6-octachlorobiphenyl in Solution No. 1
calculated versus 3,3',4,4'-tetrachlorobiphenyl-dg on a day-to-day basis.
-------
I
OJ
DATE
-c TnTfi.'.csitor.o-nminjiYL. ISOIIKR 11210
/nni:.ARI:A)/(AMT./r.EF.AIir.) IAV: 0.2571
0. 320
U.3,0 -
n.:•»
X
j
i
f j
V
>
ST.MiV.-
0.029
^ J: ST.|:EV.-
tt.239
SOU - 1.0
^•-
X
X
<
5/11/82
5/21/02
6/
Figure A-22. Relative response factor for 2,2' ,3,3 ' ,4,4' ,5,6,6' -nonachlorobiphenyl in Solution No . ', 1
calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis.
-------
>
ro
RJISrOIISE LlB:01.1t. 1IASS: 498 tEEF.OmMH. li MASS: 299)
aiP!2.2>.3.3g.4.4'.5.5i.C.6i-DLCAail.OP,0-DII'lllillYL. ISOHHIi D 209
ni-F:D-t> TEU!/\aiLOBO-niPiiwri'L. isoiiiiis 11210
(AI',i;A/r,lil;.Ai!li'.>/(AHT./IlEl:.MIT.) IAV: 0.223)
0.2/0
it. ?r.n
0.2SO
0.210
0.2JO
0.220
0.210
0.200
0. 100
UATI:
«
*
?
4
Sl.HSV.-
C.022
,? sr.otv.-
0.077
StW "1.8
5/H/82
5/2-1/U2
C/ 3/02
Figure A-23. Relative response factor for 2,2',3,3',4,4',5,5',6,6'-decachlorobiphenyl in Solution No. 1
calculated versus 3,3' ,4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis.
-------
APPENDIX B
ANALYTICAL METHOD: THE ANALYSIS OF BY-PRODUCT CHLORINATED
BIPHENYLS IN COMMERCIAL PRODUCTS AND PRODUCT WASTES
B-l
-------
THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS IN
COMMERCIAL PRODUCTS AND PRODUCT WASTES
1.0 Scope and Application
1.1 This is a gas chromatographic/electron impact mass spectrometric
(GC/EIMS) method applicable to the determination of chlorinated
biphenyls (PCBs) in commercial products and product wastes. The
PCBs present may originate either as synthetic by-products or as
contaminants derived from commercial PCB products (e.g., Aroclors).
The PCBs may be present as single isomers or complex mixtures and
may include all 209 congeners from monochlorobiphenyl through
decachlorobiphenyl listed in Table 1.
1.2 The detection and quantitation limits are dependent upon the com-
plexity of the sample matrix and the ability of the analyst to
remove interferents and properly maintain the analytical system.
The method accuracy and precision will be determined in future
studies.
1.3 This method is restricted to use by or under the supervision of
analysts experienced in the use of gas chromatography/mass spec-
trometry (GC/MS) and in the interpretation of gas chromatograms
and mass spectra. Prior to sample analysis, each analyst must
demonstrate the ability to generate acceptable results with this
method by following the procedures described in Section 14.2.
1.4 The validity of the results depends on equivalent recovery of the
analyte and 13C PCBs. If the *3C PCBs are not thoroughly incor-
porated in the matrix, the method is not applicable.
1.5 During the development and testing of this method, certain analyti-
cal parameters and equipment designs were found to affect the valid-
ity of the analytical results. Proper use of the method requires
that such parameters or designs must be used as specified. These
items are identified in the text by the word "must." Anyone wish-
ing to deviate from the method in areas so identified must demon-
strate that the deviation does not affect the validity of the data.
Alternative test procedure approval must be obtained from the
Agency. An experienced analyst may make modifications to param-
eters or equipment identified by the term "recommended." Each
time such modifications are made to the method, the analyst must
repeat the procedure in Section 14.2. In this case, formal ap-
proval is not required, but the documented data from Section 14.2
must be on file as part of the overall quality assurance program.
B-2
-------
TABLE 1. NUMBERING OF PCB CONGENERS3
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1$
17
IS
19
20
21
22
23
24
25
26
27
23
29
33
31
32
33
34
35
36
37
33
39
40
41
42
43
44
45
46
47
48
49
SO
SI
Structure
Honeehlorott
-------
2.0 Summary
2.1 The process or product must be sampled such that the specimen col-
lected for analysis is representative of the whole. Statistically
designed selection of the sampling position, time, or discrete
product units should be employed, the sample must be preserved .
to prevent PCB loss prior to analysis. Customary inventory stor-
age may be adequate for products. For intermediates, process sam-
ples, and other non-product specimens, storage at 4°C with op-
tional preservation at low pH is recommended.
2.2 The sample is mechanically homogenized and subsampled if necessary.
The sample is then spiked with four 13C PCB surrogates and the
surrogates incorporated by further mechanical agitation.
2.3 The surrogate-spiked sample is extracted and cleaned up at the
discretion of the analyst. Simple dilution or direct injection
is permissible. Possible extraction techniques include liquid-
liquid partition, thermal desorption, and sorption onto resin
columns followed by solvent desorption. Cleanup techniques may
include liquid-liquid partition, sulfuric acid cleanup, saponifi-
cation, adsorption chromatography, gel permeation chromatography,
or a combination of cleanup techniques. The sample is diluted or
concentrated to a final known volume for instrumental determina-
tion.
2.4 The PCB content of the sample extract is determined by capillary
(preferred) or packed column gas chromatography/electron impact
mass spectrometry (CGC/EIMS or PGC/EIMS) operated in the selected
ion monitoring (SIM), full scan, or limited mass scan (LMS) mode.
2.5 PCBs are identified^by comparison of their retention time and mass
spectral intensity ratios to those in calibration standards.
2.6 PCBs are quantitated against the response factors for a mixture
of 11 PCB congeners, using the response of the 13C surrogate to
compensate for losses in workup and determination and instrument
variability.
2.7 The PCBs identified by the SIM technique may be confirmed by full
scan CGC/EIMS, retention on alternate GC columns, other mass spec-
trometric techniques, infrared spectrometry, or other techniques,
provided that the sensitivity and selectivity of the technique are
demonstrated to be comparable or superior to GC/EIMS.
2.8 The analysis time is dependent on the extent of workup employed.
The time required for instrumental analysis of a single sample,
excluding data reduction and reporting, is about 30 to 45 rain.
2.9 Appropriate quality control (QC) procedures are included to assess
the performance of the analyst and estimate the quality of the re-
sults. These QC procedures include the demonstration of laboratory
capability: periodic analyst certification, the use of control
B-4
-------
charts, and the analysis of blanks, replicates, and standard addi-
tion samples. A quality assurance (QA) plan must be developed for
each laboratory.
2.10 While several options are available throughout this method, the
recommended procedure to be followed is:
2.10.1 The sample is collected according to a scheme which per-
mits extrapolation of the sample data to the whole pro-
duct or product waste.
2.10.2 The sample is preserved to prevent any loss of PCBs or
changes in matrix which may adversely affect recovery.
2.10.3 The sample is mechanically homogenized and subsampled if
necessary.
2.10.4 The sample is spiked with four 13C PCB surrogates
(4-chlorobiphenyl; 3,3',4,4'-tetrachlorobiphenyl;
2,2',3,3',5,5',6,6'-octachlorobiphenyl; and decachloro-
biphenyl).
2.10.5 Normally, the sample is extracted, although dilution may
also be used.
2.10.6 The extract is cleaned up and concentrated to an appro-
priate volume.
2.10.7 An aliquot of the extract is analyzed by CGC/EIMS oper-
ated in the SIM mode. On-eolumn injections onto a 15-m
DB-5 capillary column, programmed (for toluene solutions)
from 110° to 325°C at 10°'/min after a 2-min hold is used.
Helium at 45-cm/sec linear velocity is used as the carrier
gas.
2.10.8 PCBs are identified by retention time and mass spectral
intensities.
2.10.9 PCBs are quantitated against the response factors for a
mixture of 11 PCB congeners.
2.10.10 The total PCBs are obtained by summing the amounts for
each homolog found, and the concentration is reported
as micrograms per gram.
3.0 Interferences
3.1 Method interferences may be caused by contaminants in solvents,
reagents, glassware, and other sample processing hardware, leading
to discrete artifacts and/or elevated baselines in the total ion
current profiles. All of these materials must be routinely demon-
strated to be free from interferences by the analysis of labora-
tory reagent blanks as described in Section 14.4.
B-5
-------
3.1.1 Glassware must be scrupulously cleaned. All glassware
is cleaned as soon as possible after use by rinsing with
the last solvent used. This should be followed by deter-
gent washing with hot water and rinses with tap water and
reagent water. The glassware should then be drained dry
and heated in a muffle furnace at 400°C for 15 to 30 min.
Some thermally stable materials, such as PCBs, may not
be eliminated by this treatment. Solvent rinses with
acetone and pesticide quality hexane may be substituted
for the muffle furnace heating. Volumetric ware should
not be heated in a muffle furnace. After it is dry and
cool, glassware should be sealed and stored in a clean
environment to prevent any accumulation of dust or other
contaminants. It is stored inverted or capped with
aluminum foil.
3.1.2 The use of high purity reagents and solvents helps to
minimize interference problems. Purification of sol-
vents by distillation in all-glass systems may be re-
quired. All solvent lots must be checked for purity
prior to use.
3.2 Matrix interferences may be caused by contaminants that are coex-
tracted from the sample. The extent of matrix interferences will
vary considerably from source to source, depending upon the nature
and diversity of the sources of samples.
4.0 Safety
4.1 The toxicity or carcinogenicity of each reagent used in this
method has not been precisely defined; however, each chemical
compound should be treated as a potential health hazard. From
this viewpoint, exposure to these chemicals must be reduced to
the lowest possible level by whatever means available. The la-
boratory is responsible for maintaining a current awareness file
of OSHA regulations regarding the safe handling of the chemicals
specified in this method. A reference file of material data han-
dling sheets should also be made available to all personnel in-
volved in the chemical analysis.
4.2 Polychlorinated biphenyls have been tentatively classified as
known or suspected human or mammalian carcinogens. Primary stan-
dards of these toxic compounds should be prepared in a hood.
Personnel must wear protective equipment, including gloves and
safety glasses.
Congeners highly substituted at the meta and para positions and
unsubstituted at the ortho positions are reported to be the most
toxic. Extreme caution should be taken when handling these com-
pounds neat or in concentrated solutions. This class includes
3,3',4,4'-tetrachlorobiphenyl (both natural abundance and isotop-
ically labeled).
B-6
-------
4.3 Diethyl ether should be monitored regularly to determine the per-
oxide content. Under no circumstances should diethyl ether be
used with a peroxide content in excess of 50 ppm, as an explosion
could result. Peroxide test strips manufactured by EM Labora-
tories (available from Scientific Products Company, Cat. No.
P1126-8 and other suppliers) are recommended for this test. Pro-
cedures for removal of peroxides from diethyl ether are included
in the instructions supplied with the peroxide test kit.
4.4 Waste disposal must be in accordance with RCRA and applicable
state rules.
5.0 Apparatus and Materials
5.1 Sampling containers - Amber glass bottles, 1-liter or other ap-
propriate volume, fitted with screw caps lined with Teflon.
Cleaned foil may be substituted for Teflon if the sample is not
corrosive. If amber bottles are not available, samples should
be protected from light using foil or a light-tight outer con-
tainer. The bottle must be washed, rinsed with acetone or methy-
lene chloride, and dried before use to minimize contamination.
5.2 Glassware - All specifications are suggestions only. Catalog
numbers are included for illustration only.
5.2.1 Volumetric flasks - Assorted sizes.
5.2.2 Pipets - Assorted sizes, Mohr delivery.
5.2.3 Micro syringes - 10.0 (Jl for packed column GC analysis,
1.0 (Jl for on-column GC analysis.
5.2.4 Chromatographic column - Chromaflex, 400 mm long x 19 mm
ID (Kontes K-420540-9011 or equivalent).
5.2.5 Gel permeation chromatograph - GPC Autoprep 1002 (An-
alytical Bio Chemistry Laboratories, Inc.) or equivalent.
5.2.6 Kuderna-Danish Evaporative Concentrator Apparatus
5.2.6.1 Concentrator tube - 10 ml, graduated (Kontes
K-570050-1025 or equivalent). Calibration must
be checked. Ground glass stopper size (519/22
joint) is used to prevent evaporation of solvent.
5.2.6.2 Evaporative flask - 500 ml (Kontes K-57001-0500
or equivalent). Attached to concentrator tube
with springs (Kontes K-662750-0012 or equiva-
lent) .
5.2.6.3 Snyder column - Three ball macro (Kontes
K-503000-0121 or equivalent).
/
B-7
-------
5.3 Balance - Analytical, capable of accurately weighing 0.0001 g.
5.4 Gas chromatography/raass spectrometer system.
5.4.1 Gas chromatograph - An analytical system complete with a
temperature programmable gas chromatograph and all re-
quired accessories including syringes, analytical columns,
and gases. The injection port must be designed for on-
column injection when using capillary columns or packed
columns. Other capillary injection techniques (split,
splitless, "Grob," etc.) may be used provided the per-
formance specifications stated in Section 7.1 are met.
5.4.2 Capillary GC column - A 12-20 m long x 0.25 mm ID fused
silica column with a 0.25 |Jm thick DB-5 bonded silicone
liquid phase (J&W Scientific) is recommended. Alternate
liquid phases may include OV-101, SP-2100, Apiezon L,
Dexsil 300, or other liquid phases which meet the perfor-
mance specifications stated in Section 7.1.
5.4.3 Packed GC column - A 180 cm x 0.2 cm ID glass column
packed with 3% SP-2250 on 100/120 mesh Supelcoport or
equivalent is recommended. Other liquid phases which
meet the performance specifications stated in Section 7.1
may be substituted.
5.4.4 Mass spectrometer - Must be capable of scanning from 150
to 550 daltons every 1.5 sec or less, collecting at least
five spectra per chromatographic peak, utilizing a 70-eV
(nominal) electron energy in the electron impact ioniza-
tion mode and producing a mass spectrum which meets all
the criteria in Table 2 when 50 ng of decafluorotriphenyl
phosphine [DFTPP, bis(perfluorophenyl)phenyl phosphine]
is injected through the GC inlet. Any GC-to-MS interface
that gives acceptable calibration points at 10 ng per
injection for each PCB isomer in the calibration stan-
dard and achieves all acceptable performance criteria
(Section 10) may be used. Direct coupling of the fused
silica column to the MS is recommended. Alternatively,
GC-to-MS interfaces constructed of all glass or glass-
lined materials are recommended. Glass can be deacti-
vated by silanizing with dichlorodimethylsilane.
5.4.5 A computer system that allows the continuous acquisition
and storage on machine-readable media of all mass spectra
obtained throughout the duration of the chromatographic
program must be interfaced to the mass spectrometer.
The data system must have the capability of integrating
the abundances of the selected ions between specified
limits and relating integrated abundances to concentra-
tions using the calibration procedures described in this
method. The computer must have software that allows
B-8
-------
TABLE 2. DFTPP KEY IONS AND ION ABUNDANCE CRITERIA
Mass Ion abundance criteria
197 Less than 1% of mass 198
198 100% relative abundance
199 5-9% of mass 198
275 10-30% of mass 198
365 Greater than 1% of mass 198
441 Present, but less than mass 443
442 Greater than 40% of mass 198
443 17-23% of mass 442
B-9
-------
searching any GC/MS data file for ions of a specific mass
and plotting such ion abundances versus time or scan
number to yield an extracted ion current profile (EICP).
Software must also be available that allows integrating
the abundance in any EICP between specified time or scan
number limits.
6.0 Reagents
6.1 Solvents - All solvents must be pesticide residue analysis grade.
New lots should be checked for purity by concentrating an aliquot
by at least as much as is used in the procedure.
6.2 Calibration standard congeners - Standards of the PCB congeners
listed in Table 3 are available from Ultra Scientific, Hope,
Rhode Island; or Analabs, North Haven, Connecticut.
6.3 Calibration standard stock solutions - Primary dilutions of each
of the individual PCBs listed in Table 3 are prepared by weighing
approximately 1-10 mg of material within 1% precision. The PCB
is then dissolved and diluted to 1.0 ml with hexane. The concen-
tration is calculated in mg/ml. The primary dilutions are stored
at 4°C in screw-cap vials with Teflon cap liners. The meniscus
is marked on the vial wall to monitor solvent evaporation. Pri-
mary dilutions are stable indefinitely if the seals are maintained.
The validity of primary and secondary dilutions must be monitored
on a quarterly basis by analyzing four quality control check sam-
ples (see Section 14.2).
6.4 Working calibration standards - Working calibration standards are
prepared that are similar in PCB composition and concentration to
the samples by mixing and diluting the individual standard stock
solutions. Example calibration solutions are shown in Table 3.
The mixture is diluted to volume with pesticide residue analysis
quality hexane. The concentration is calculated in ng/ml as the
individual PCBs. Dilutions are stored at 4°C in narrow-mouth,
screw-cap vials with Teflon cap liners. The meniscus is marked
on the vial wall to monitor solvent evaporation. These secondary
dilutions can be stored indefinitely if the seals are maintained.
These solutions are designated "CSxxx," where the xxx is used to
encode the nominal concentration in ng/ml.
6.5 Alternatively, certified stock solutions similar to those listed
in Table 3 may be available from a supplier, in lieu of the pro-
cedure described in Section 6.4.
6.6 DFTPP standard - A 50-ng/Ml solution of DFTPP is prepared in ace-
tone or another appropriate solvent.
6.7 Surrogate standard stock solution - The four 13C-labeled PCBs
listed in Table 4 may be available from a supplier as a certi-
fied solution. This solution may be used as received or diluted
further. These solutions are designated "SSxxx," where the xxx
is used to encode the nominal concentration in [Jg/ml.
B-10
-------
TABLE 3. CONCENTRATIONS OF CONGENERS IN PCB CALIBRATION STANDARDS (rig/ml)a
Homolog
1
1
2
3
4
5
6
7
8
9
10
4
1
4
8
10
Congener
no.
1
3
7
30
50
97
143
183
202
207
209
210 (IS)
211 (RS)
212 (RS)
213 (RS)
214 (RS)
CS1000
1,040
1,000
1,040
1,040
1,520
1,740
1,920
2,600
4,640
5,060
4,240
255
104
257
\ 407
502
CS100
104
100
104
104
152
174
192
260
464
506
424
255
104
257
407
502
CS050
52
50
52
52
76
87
96
130
232
253
212
255
104
257
407
502
CS010
10
10
10
10
15
17
19
26
46
51
42
255
104
257
407
502
a Concentrations given as examples only.
B-ll
-------
TABLE 4. COMPOSITION OF SURROGATE SPIKING SOLUTION (SS100) CONTAINING
13C-LABELED PCBsa
Congener
no.
211
212
213
214
Compound
(!' ,2' ,3' ,4' ,5' ,6'-13C6)4-chlorobiphenyl
(13Ci2)3,3',4,4' -tetrachlorobiphenyl
(13C12)2,2! ,3,3' ,5,5* ,6,6'-octachlorobiphenyl
(13C12)decachloirobiphenyl
Concentration
(Mg/ml)
104
257
395
502
a Concentrations given as examples only.
B-12
-------
6.8 Internal standard solution - A solution of de-3,3*,4,4'-tetra-
chlorobiphenyl is prepared at a nominal concentration of 1-10
mg/ml in hexane. The solution is further diluted to give a work-
ing standard.
6.9 Solution stability - The calibration standard, surrogate, and
DFTPP solutions should be checked frequently for stability. These
solutions should be replaced after 6 months, or sooner if compari-
son with quality control check samples indicates compound degrada-
tion or concentration change.
6.10 Quality control check samples will be supplied by the Agency.
7.0 Calibration
7.1 The gas chromatograph must meet the minimum operating parameters
shown in Tables 5 and 6, daily. If all criteria are not met, the
analyst must adjust conditions and repeat the test until all cri-
teria are met.
7.2 The mass spectrometer jmust meet the minimum operating parameters
shown in Tables 2, 7, and 8, daily. If all criteria are not met,
the analyst must retune the spectrometer and repeat the test un-
til all conditions are met.
7.3 The PCB response factors (RF ) must be determined using Equation
7-1 for the analyte homologs?
An x M.
RF = -r* ^ Eq. 7-1
p A. x M ^
is p
where RF = response factor of a given PCB congener
A = area of the characteristic ion for the PCB congener
peak
M = mass of PCB congener injected (nanograms)
A. = area of the characteristic ion for the internal
standard peak
M. = mass of internal standard injected (nanograms)
IS
Using the same conditions as for RF , the surrogate response
factors (RF ) must be determined using Equation 7-2.
s
A x M.
RF = ^ £5. Eq. 7-2
s A. x M H
is s
where A = area of the characteristic ion for the surrogate peak
S
M = mass of surrogate injected (nanograms)
S
Other terms are the same as defined in Equation 7-1.
B-13
-------
TABLE 5. OPERATING PARAMETERS FOR CAPILLARY COLUMN GAS CHROMATOGRAPHIC SYSTEM
Parameter
Recommended
Tolerance
Gas chromatograph
Column
Liquid phase
Liquid phase thickness
Carrier gas
Carrier gas velocity
Injector
Injector, temperature
Injection volume
Initial column temperature
Column temperature program
Finnigan 9610
15 m x 0.255 mm ID
Fused silica
DB-5 (J&W)
0.25 |Jm
Helium
45 cm/sec
f+
On-column (J&W)
Optimum performance
1.0 [Jlc
70°C (2 min)d
70°-325°C at 10°C/mine
Other
Other
Other nonpolar
or semipolar
< 1 |Jm
Hydrogen
Optimum performance
Other
Optimum performance
Other
Other
Other
Separator
Transfer line temperature
Tailing factor
Peak width1
None
280°C
0.7-1.5
7-10 sec
Glass jet or othe
Optimum*
0.4-3
< 15 sec
a Substitutions permitted with any common apparatus or technique provided
performance criteria are met.
b Measured by injection of air or methane at 270°C oven temperature.
c For on-column injection, manufacturer's instructions should be followed
regarding injection technique.
d With on-column injection, initial temperature equals boiling point of the
solvent; in this instance, hexane.
e C12C11Q elutes at 270°C. Programming above this temperature ensures a
clean column and lower background on subsequent runs.
f Fused silica columns may be routed directly into the ion source to pre-
vent separator discrimination and losses.
g High enough to elute all PCBs, but not high enough to degrade the column
if routed through the transfer line.
h Tailing factor is width of front half of peak at 10% height divided by width
of back half of peak at 10% height for single PCB congeners in solution CSxxx,
i Peak width at 10% height for a single PCB congener is CSxxx.
B-14
-------
TABLE 6. OPERATING PARAMETERS FOR PACKED COLUMN GAS CHROMATOGRAPHY SYSTEM
Parameter
Recommended
Tolerance
Gas chromatograph
Column
Finnigan 9610
180 cm x 0.2 cm ID
Other3
Other
Column packing
Carrier gas
Carrier gas flow rate
Injector
Injector temperature
Injection volume
Initial column temperature
Column temperature program
Separator
Transfer line temperature
Tailing factor
Peak width
glass
31 SP-2250 on 100/
120 mesh Supelcoport
Helium
30 ml/tnin
On-column
250°C
1.0 Ml
150°C, 4 min
150°-260°C at 8°/min
Glass jet
280°C
0.7-1.5
10-20 sec
Other nonpolar
or semipolar
Hydrogen
Optimum performance
Other
^ . b
Optimum
^ 5 Ml
Other
Other
Other
ft
Optimum
0.4-3
< 30 sec
a Substitutions permitted if performance criteria are met,
b High enough to elute all PCBs.
c Tailing factor is width of front half of peak at 10% height divided by
width of back half of peak at 10% height for single PCB congeners in solu-
tion CSxxx.
d Peak width at 10% height for a single PCB congener in CSxxx.
B-15
-------
TABLE 7. OPERATING PARAMETERS FOR QUADRUPOLE MASS SPECTROMETER SYSTEM
Parameter Recommended Tolerance
Mass spectrometer
Data system
Scan range
Scan time
Resolution
Ion source temperature
Electron energy
Trap current
Multiplier voltage
Preamplifier sensitivity
Finnigan 4023
Incos 2400
95-550
1 sec
Unit
280°C
70 eV
0.2 mA
-1,600 V
10"6 A/V
Other3
Other
Other
Otherb
Optimum performance
200°-300°C
Optimum performance
Optimum performance
Optimum performance
Set for desired
working range
a Substitutions permitted if performance criteria are met.
b Greater than five data points over a GC peak is a minimum.
c Filaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no separator is used.
B-16
-------
TABLE 8. OPERATING PARAMETERS FOR MAGNETIC SECTOR MASS SPECTROMETER SYSTEM
Parameter
Mass spectrometer
Data system
Scan range
Scan mode
Cycle time
Resolution
Ion source temperature
Electron energy
Emission current
Filament current
Multiplier
Recommended
Finnigan MAT 311A
Incos 2400
98-550
Exponential
1.2 sec
1,000
280°C
70 eV
1-2 mA
Optimum
-1,600 V
Tolerance
Other3
Other
Other
Other
Otherb
> 500
250°-300°C
70 eV
Optimum
Optimum
Optimum
a Substitutions permitted if performance criteria are met.
b Greater than five data points over a GC peak is a minimum.
c Filaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no separator is used.
B-17
-------
If specific congeners are known to be present and if standards
are available, selected RF values may be employed. For general
samples, solutions CSxxx and SSxxx or a mixture (Tables 3 and 4),
with a similar level of internal standard (dg-3,31,4,4'-tetra-
chlorobiphenyl) added, may be used as the response factor solution.
The PCB-surrogate pairs to be used in the RF calculation are listed
in Table 9.
Generally, only the primary ions of both the analyte and surrogate
are used to determine the RF values. If alternate ions are to be
used in the quantitation, the RF must be determined using that
characteristic ion.
The RF value must be determined in a manner to assure ±20% accu-
racy and precision. For instruments with good day-to-day preci-
sion, a running mean (RF) based on seven values determined once
each day may be appropriate. Other options include, but are not
limited to, triplicate determinations of a single concentration
spaced throughout a day or determination of the RF at three dif-
ferent levels to establish a working curve.
If replicate RF values differ by greater than ±10% RSD, the system
performance should be monitored closely. If the RSD is greater
than ±20%, the data set must be considered invalid and the RF re-
determined before further analyses are done.
7.4 If the GC/EIMS system has not been demonstrated to yield a linear
response or if the analyte concentrations are more than two orders
of magnitude different from those in the RF solution, a calibration
curve must be prepared. If the analyte and RF solution concentra-
tions differ by more than one order of magnitude, a calibration
curve should be prepared. A calibration curve should be estab-
lished with triplicate determinations at three or more concentra-
tions bracketing the analyte levels.
7.5 The relative retention time (RRT) windows for the 10 homologs and
surrogates must be determined. If all congeners are not available,
a mixture of available congeners or an Aroclor mixture (e.g.,
1016/1254/1260) may be used to estimate the windows. The windows
must be set wider than observed if all isomers are not determined.
Typical RRT windows for one column are listed in Table 10. The
windows may differ substantially if other GC parameters are used.
8.0 Sample Collection, Handling, and Preservation
8.1 Amber glass sample containers should have Teflon-lined screw caps.
With noncorrosive samples, methylene chloride-washed aluminum foil
liners may be substituted. The volume and configuration are deter-
mined by the amount of sample to be collected and its physical
properties. For dry powders, other containers such as heavy-walled
polyethylene bags may be appropriate.
B-18
-------
TABLE 9. PAIRINGS OF ANALYTE, CALIBRATION. AND SURROGATE COMPOUNDS
Analyte
Congener
no.
1
2,3
4-15
16-39
40-81
82-127
128-169
170-193
194-205
206-208
209
Compound
Calibration standard
Congener
no.
2~Ci2H9Cl 1
3- and 4-C12H9Cl 3
C12H7C13
C12H6C14
C12H5C15
C12H4C16
C12H2Clg
C12HClg
30
50
97
143
183
202
207
209
2
4
2,4
2,4,
2,2'
2,2'
2,2'
2,2'
2,2'
2,2'
p r>
L12L
6
,4
,3
,3
,3
,3
,3
li
Compound
,6
',4,
,4,5
',4,
,3',
3'
, j ,
0
5
,6'
4', 5', 6
5, 5', 6, 6'
4, 4', 5, 6, 6'
Surrogate
Congener
no.
211
211
211
212
212
212
212
213
213
213
214
Compound
13C6-4
J3C6-4
13C12-3,3'
13C12-3,3'
13C12-3,3'
13C12-3,3'
13C12-2,2'
13C12-2,2'
13C12-2,2'
13Ci2Cl10
,4
,4
,4
,4
,3
,3
,3
,4'
,4'
,4'
,4'
,3',5,5'
,3', 5, 5'
3 ' 55*
,6,6'
,6,6'
,6,6'
a Ballschmiter numbering system, see Table 1.
-------
TABLE 10. RELATIVE RETENTION TIME (RRT) RANGES OF PCB HOMOLOGS
VERSUS d6-3.3',4,4'-TETRACHLOROBIPHENYL
PCB
homolog
Monochloro
Dichloro
Trichloro
Tetrachloro
Pentachloro
Hexachloro
Heptachloro
Octachloro
Nonachloro
Decachloro
No. of
isomers
measured
3
10
9
16
12
13
4
6
3
1
Observed range
of RRTsa
0.40-0.50
0.52-0.69
0.62-0.79
0.72-1.01
0.82-1.08
0.93-1.20
1.09-1.30
1.19-1.36
1.31-1.42
1.44-1.45
Calibration
Congener
no.
1
3
7
30
50
97
143
183
202
207
209
solution
Observed
RRTa
0.43
0.50
0.58
0.65
0.75
0.98
1.05
1.15
1.19
1.33
1.44
Projected
range of
RRTs
0.35-0.55
0.35-0.80
0.35-1.10
0.55-1.05
0.80-1.10
0.90-1.25
1.05-1.35
1.10-1.50
1.25-1.50
1.35-1.50
a The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were
measured versus 3,3',4,4'-tetrachlorobiphenyl-dg (internal standard) using
a 15-m J&W DB-5 fused silica column with a temperature program of 110°C
for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an on-
column injector. A Finnigan 4023 Incos quadrupole mass spectrometer oper-
ating with a scan range of 95-550 daltons was used to detect each PCB
congener.
b The projected relative retention windows account for overlap of eluting
homologs and take into consideration differences in operating systems
and lack of all possible 209 PCB congeners.
B-20
-------
8.2 Sample bottle preparation
8.2.1 All sample containers and caps should be washed in deter-
gent solution, rinsed with tap water, and then with dis-
tilled water. The bottles and caps are allowed to drain
dry in a contaminant-free area. Then the caps are rinsed
with pesticide grade hexane and allowed to air dry.
8.2.2 Sample bottles are heated to 400°C for 15 to 20 min or
rinsed with pesticide grade acetone or hexane and allowed
to air dry.
8.2.3 The clean bottles are stored inverted or sealed until use.
8.3 Sample collection
8.3.1 The primary consideration in sample collection is that
the sample collected be representative of the whole.
Therefore, sampling plans or protocols for each individ-
ual producer's situation will have to be developed. The
recommendations presented here describe general situa-
tions. The number of replicates and sampling frequency
also must be planned prior to sampling.
8.3.2 Discrete product units - If the product is small enough
that one or more discrete units would be used as the an-
alytical sample, a statistically random sampling approach
is recommended.
8.3.3 Liquids or free-flowing solids - If possible, the source
is mixed thoroughly before collecting the sample. If
mixing is impractical, the sample should be collected
from a representative area of the source. If the liquid
is flowing through an enclosed system, sampling through
a valve should be randomly timed.
8.3.4 Solids - Larger bulk solids which must be subsampled to
get a reasonably sized analytical sample must be treated
on a case-by-case basis. A representative sample should
be obtained by designing a sampling location selection
scheme such that all parts of the whole have a finite,
known probability of inclusion. Based on such a scheme,
the PCB content of the sample can be used to extrapolate
to the content of the whole.
8.4 Sample preservation - Product samples should be stored as the bulk
or packaged product inventory would be stored, or in a cool, dry,
dark area. Intermediates, process samples, or other non-product
specimens should be stored at 4°C. If there is a possibility of
microbial degradation, addition of H2S04 during collection to a
pH < 2 is recommended. A test strip is used to monitor pH. Stor-
age times in excess of 4 weeks are not recommended.
B-21
-------
If residual chlorine is present in the sample, it should be
quenched with sodium thiosulfate. EPA Methods 330.4 and 330.5
may be used to measure the residual chlorine.1 Field test kits
are available for this purpose.
9.0 Sample Preparation
Since a wide variety of matrices may be subjected to analysis by this
method, the extraction/cleanup procedure cannot be specified. This
section describes general guidelines for subsampling, addition of 13C
surrogates, dilution, extraction, cleanup, extract concentration, and
other sample preparation procedures.
9.1 Sample homogenization and subsarapling - The sample is homogenized
by shaking, blending, shredding, crushing, or other appropriate
mechanical technique. A representative subsample of 100 g or other
known mass is then taken. The sample size is dependent upon the
anticipated PCB levels and difficulty of the subsequent extraction/
cleanup steps.
Note: The precision of the mass determination at this step will
be reflected in the overall method precision. Therefore, an an-
alytical balance must be used to assure that the weight is accu-
rate to ±1% or better.
9.2 Surrogate addition - An appropriate volume of surrogate solution
SSxxx is pipetted into the sample. The final concentration of the
surrogates must be in the working range of the calibration and
well above the matrix background. The surrogates are thoroughly
incorporated by further mechanical agitation. For nonviscous
liquids, shaking for 30 sec should be sufficient. For viscous
liquids or free-flowing solids, 10-min tumbling is recommended.
In cases where inadequate incorporation may be expected, such as
solids, overnight equilibration with agitation is recommended.
Note: The volume measurement of the spiking solution is critical
to the overall method precision. The analyst must exercise cau-
tion that the volume is known to ±J% or better. Where necessary,
calibration of the pipet is recommended.
9.3 Sample preparation (extraction/cleanup) - After addition of the
surrogates, the sample is further treated at the discretion of
the analyst, provided that the GC/EIMS response of the four sur-
rogates meets the criteria listed in Section 7.0. The literature
pertaining to these techniques has been reviewed.2 Several pos-
sible techniques are presented below for guidance only. The ap-
plicability of any of these techniques to a specific sample ma-
trix must be determined by the precision and accuracy of the 13C
PCB surrogate recoveries, as discussed in Section 14.2.
B-22
-------
9.3.1 Extraction1
9.3.1.1 Dilution - In some cases, where the PCB concen-
tration is high, a simple volumetric dilution
with an appropriate solvent may be sufficient
sample preparation.
9.3.1.2 Direct injection - If sample viscosity permits,
direct injection with no dilution is permissible.
9.3.1.3 Liquid-liquid extraction - If the matrix is
aqueous (or another solvent in which PCBs are
only slightly soluble), a liquid-liquid parti-
tion may be effective. The solvent, number of
extractions, solvent-to-sample ratio, and other
parameters are chosen at the analyst's discretion.
9.3.1.4 Sorbent column extraction - PCBs may be isolated
from free-flowing liquids onto sorbent columns.
The selection of sorbent (XAD, Porapak, carbon-
polyurethane foam, etc.) will depend on the na-
ture of the matrix. The available methods have
been reviewed.2
9.3.1.5 Thermal desorption - If the matrix is nonvol-
atile, thermal desorption of the PCBs onto a
sorbent column, filter, or cold trap may be an
effective extraction/cleanup method.
9.3.2 Cleanup - Several tested cleanup techniques are described
below. All but the base cleanup (9.3.2.8) were previously
validated for PCBs in transformer fluids.3 Depending
upon the complexity of the sample, one or more of the
techniques may be required to fractionate the PCBs from
interferences. For most cleanups a concentrated (1-5 ml)
extract should be used.
9-3.2.1 Acid cleanup
9.3.2.1.1 Place 5 ml of concentrated sulfuric
acid into a 40-ml narrow-mouth screw-
cap bottle. Add the sample extract.
Seal the bottle with a Teflon-lined
screw cap and shake for 1 min.
9.3.2.1.2 Allow the phases to separate, transfer
the sample (upper phase) with three
rinses of 1-2 ml solvent to a clean
container and concentrate to an ap-
propriate volume.
B-23
-------
9.3.2.1.3 Analyze as described in Section 10.0.
9.3.2.1.4 If the sample is highly contaminated,
a second or third acid cleanup may
be employed.
9.3.2.2 Florisil column cleanup
9.3.2.2.1 Variations among batches of Florisil
(PR grade or equivalent) may affect
the elution volume of the various
PCBs. For this reason, the volume
of solvent required to completely
elute all PCBs must be verified by
the analyst. The weight of Florisil
can then be adjusted accordingly.
9.3.2.2.2 Place a 20-g charge of Florisil,
activated overnight at 130°C, into a
Chromaflex column. Settle the Flor-
isil by tapping the column. Add about
1 cm of anhydrous sodium sulfate to
the top of the Florisil. Pre-elute
the column with 70-80 ml of hexane.
Just before the exposure of the sodium
sulfate layer to air, stop the flow.
Discard the eluate.
9.3.2.2.3 Add the sample extract to the column.
9.3.2.2.4 Carefully wash down the inner wall
of the column with 5 ml of hexane.
9.3.2.2.5 Add 200 ml of 6% ethyl ether/hexane
and set the flow to about 5 ml/min.
9.3.2.2.6 Collect 200 ml of eluate in a Kuderna-
Danish flask. All the PCBs should be
in this fraction. Concentrate to an
appropriate volume.
9.3.2.2.7 Analyze the sample as described in
Section 10.0.
9.3.2.3 Alumina column cleanup
9.3.2.3.1 Adjust the activity of the alumina
(Fisher A450 or equivalent) by heat-
ing to 200°C for 2 to 4 hr. When
cool, add 3% water (wt:wt) and mix
until uniform. Store in a tightly
sealed bottle. Allow the deactivated
alumina to equilibrate at least 1/2 hr
before use. Reactivate weekly.
B-24
-------
9.3.2.3.2 Variations between batches of alumina
may affect the elution volume of the
various PCBs. For this reason, the
volume of solvent required to com-
pletely elute all of the PCBs must
be verified by the analyst. The
weight of alumina can then be ad-
justed accordingly.
9.3.2.3.3 Place a 50-g charge of alumina into
a Chromaflex column. Settle the
alumina by tapping. Add about 1 cm
of anhydrous sodium sulfate. Pre-
elute the column with 70-80 ml of
hexane. Just before exposure of the
sodium sulfate layer to air, stop
the flow. Discard the eluate.
9.3.2.3.4 Add the sample extract to the column.
9.3.2.3.5 Carefully wash down the inner wall
of the column with 5 ml of hexane.
9.3.2.3.6 Add 295 ml of hexane to the column.
9.3.2.3.7 Discard the first 50 ml.
9.3.2.3.8 Collect 250 ml of the hexane in a
Kuderna-Danish flask. All of the
PCBs should be in this fraction.
Concentrate to an appropriate volume.
9.3.2.3.9 Analyze the sample as described in
Section 10.0.
9.3.2.4 Silica gel column cleanup
9.3.2.4.1 Activate silica gel (Davison Grade
950 or equivalent) at 135°C overnight.
9.3.2.4.2 Variations between batches of silica
gel may affect the elution volume of
the various PCBs. For this reason,
the volume of solvent required to
completely elute all of the PCBs must
be verified by the analyst. The
weight of silica gel can then be ad-
justed accordingly.
B-25
-------
9.3.2.4.3 Place a 25-g charge of activated
silica gel into a Chromaflex column.
Settle the silica gel by tapping the
column. Add about 1 cm of anhydrous
sodium sulfate to the top of the
silica gel.
9.3.2.4.4 Pre-elute the column with 70-80 ml
of hexane. Discard the eluate. Just
before exposing the sodium sulfate
layer to air, stop the flow.
9.3.2.4.5 Add the sample extract to the column.
9.3.2.4.6 Wash down the inner wall of the column
with 5 ml of hexane.
9.3.2.4.7 Elute the PCBs with 195 ml of 10%
diethyl ether in hexane (v:v).
9.3.2.4.8 Collect 200 ml of the eluate in a
Kuderna-Danish flask. All of the
PCBs should be in this fraction.
Concentrate to an appropriate volume.
9.3.2.4.9 Analyze the sample as described in
Section 10.0.
9.3.2.5 Gel permeation cleanup
9.3.2.5.1 Set up and calibrate the gel permea-
tion chromatograph with an SX-3 column
according to the Autoprep instruction
manual. Use 15% methylene chloride
in cyclohexane (v:v) as the mobile
phase.
9.3.2.5.2 Inject 5.0 ml of the sample extract
into the instrument. Collect the
fraction containing the PCBs (see
Autoprep operator's manual) in a
Kuderna-Danish flask equipped with
a 10-ml ampul.
9.3.2.5.3 Concentrate the PCB fraction to an
appropriate volume.
9.3.2.5.4 Analyze the sample as described in
Section 10.0.
B-26
-------
9.3.2.6 Acetonitrile partition
9.3.2.6.1 Place the sample extract into a 125-ml
separately funnel with enough hexane
to bring the final volume to 15 ml.
Extract the sample four times by shak-
ing vigorously for 1 min with 30-ml
portions of hexane-saturated acetoni-
trile.
9.3.2.6.2 Combine and transfer the acetonitrile
phases to a 1-liter separatory funnel
and add 650 ml of distilled water
and 40 ml of saturated sodium chloride
solution. Mix thoroughly for about
30 sec. Extract with two 100-ml por-
tions of hexane by vigorously shaking
about 15 sec.
9.3.2.6.3 Combine the hexane extracts in a
1-liter separatory funnel and wash
with two 100-ml portions of distilled
water. Discard the water layer and
pour the hexane layer through an 8-10
cm anhydrous sodium sulfate column
into a 500-ml Kuderna-Danish flask
equipped with a 10-ml ampul. Rinse
the separatory funnel and column with
three 10-ml portions of hexane.
9.3.2.6.4 Concentrate the extracts to an ap-
propriate volume.
9.3.2.6.5 Analyze as described in Section 10.0.
9.3.2.7 Florisil slurry cleanup
9.3.2.7.1 Place the sample extract into a 20-ml
narrow-mouth screw-cap container.
Add 0.25 g of Florisil (PR grade or
equivalent). Seal with a Teflon-lined
screw cap and shake for 1 min.
9.3.2.7.2 Allow the Florisil to settle; then
decant the treated solution into a
second container with rinsing. Con-
centrate the sample to an appropriate
volume. Analyze as described in Sec-
tion 10.0.
B-27
-------
9.3.2.8 Base cleanup4
9.3.2.8.1 Quantitatively transfer the concen-
trated extract to a 125-ral extraction
flask with the aid of several small
portions of solvent.
9.3.2.8.2 Evaporate the extract just to dryness
with a gentle stream of dry filtered
nitrogen, and add 25 ml of 2.5% alco-
holic KOH.
9.3.2.8.3 Add a boiling chip, put a water con-
denser in place, and allow the solu-
tion to reflux on a hot plate for 45
min.
9.8.2.8.4 After cooling, transfer the solution
to a 250-ml separatory funnel with
25 ml of distilled water.
9.3.2.8.5 Rinse the extraction flask with 25
ml of hexane and add it to the
separatory funnel.
9.3.2.8.6 Stopper the separatory funnel and
shake vigorously for at least 1 min.
Allow the layers to separate, and
transfer the lower aqueous phase to
a second separatory funnel.
9.3.2.8.7 Extract the saponification solution
with a second 25-ml portion of hexane.
After the layers have separated, add
the first hexane extract to the sec-
ond separatory funnel and transfer
the aqueous alcohol layer to the
original separatory funnel.
9.3.2.8.8 Repeat the extraction with a third
25-ml portion of hexane. Discard
the saponification solution, and com-
bine the hexane extracts.
9.3.2.8.9 Concentrate the hexane layer to an
appropriate volume, and analyze as
described in Section 10.0.
B-28
-------
10.0 Gas Chromatographic/Electrop Impact Mass Spectrometric Determination
10.1 Internal standard addition - An appropriate volume of the internal
standard solution is pipetted into the sample. The final concen-
tration of the internal standard must be in the working range of
the calibration and well above the matrix background. The inter-
nal standard is thoroughly incorporated by mechanical agitation.
Note: The volumetric measurement of the internal standard solu-
tion is critical to the overall method precision. The analyst
must exercise caution that the volume is known to be ±1% or better.
Where necessary, calibration of the pipet is recommended.
10.2 Tables 2, and 5 through 8 summarize the recommended operating con-
ditions for analysis. Figure 1 presents an example of a chromato-
gram.
10.3 While the highest available chromatographic resolution is not a
necessary objective of this protocol, good chromatographic per-
formance is recommended. With the high resolution of CGC, the
probability that the chromatographic peaks consist of single com-
pounds is higher than with PGC. Thus, qualitative and quantita-
tive data reduction should be more reliable.
10.4 After performance of the system has been certified for the day
and all instrument conditions set according to Tables 2, and 5
through 8, inject an aliquot of the sample onto the GC column.
If the response for any ion, including surrogates and internal
standards, exceeds the working range of the system, dilute the
sample and reanalyze. If the responses of surrogates, internal
standards, or analytes are below the working range, recheck the
system performance. If necessary, concentrate the sample and re-
analyze .
*
10.5 Record all data on a digital storage device (magnetic disk, tape,
etc.) for qualitative and quantitative data reduction as discussed
below.
11.0 Qualitative Identification
11.1 Selected ion monitoring (SIM) or limited mass scan (IMS) data -
The identification of a compound as a given PCB homolog requires
that two criteria be met:
11.1.1 (1) The peak must elute within the retention time window
set for that homolog (Section 7.5); and (2) the ratio of
two ions obtained by SIM (Table 11) or by IMS (Table 12)
must match the natural ratio within ±20%. The analyst
must search the higher mass windows, in particular M+70,
to prevent misidentification of a PCB fragment ion cluster
as the parent.
B-29
-------
100.(I-,
nic_
w
i
UJ
o
c
O)
CX
•H
O
o
u
o
d
o
CM
o
CO
c c
01 01
rC f.
cx ex
•H -H
O O OJ
t-t r-l 4-1
O O n)
rH rH 60
,C ^ O
U 0 VJ
O O M
C C p
o o to
& & ^
C7'J
c
Ol
ex
•H
.£>
O
r-l
O
rH
,£
O
•H
P
C
01
cx
•H
.a
o
t-i
o
M
H
T T
_.L_JI Lm.
c
O)
p.
o
r-l
O
rH
^2
O
cfl
t-j
4J
Ol
H
I
vO
CM
CM
U7-I
1 J^.
10:03 ,
-a
U)
r-t
n)
c
r-l
QJ
4-J
c
•rH
" —
\o
•a
rH
c
O)
cx
•H
,0
O
r-^lf"
rH
0
Cfl
r-4
4J
01
H
1
-
rH -
rC *•
CX CO
•H
,£> CO
O
^
O
rH
f.
0
CO
4-1
C
01
PH
1
tn
"*-
co
™ (
1
01
4-1
Cfl
60
O
1-1
r-l
(fl
v — '
u
co
rH
1
rH
C
01
ex
•H
c\
O
V't*
rH
U
CO rH
r-l >*»
4-j n
11 0)
H ,C
1 CX
•H
-.
•.
m
r.
—
~^
O
M
O
O
CO
O
Ol
Q
O
co
C
Ol
cx ^.
•H Ol
,0 4J
O CO
r-l 60
O O
rH t-l
f. r-l
O 3
cfl W
O '"S
O)
O
2.3:20
SCA'l
Tlllii
Figure 1. Capillary gas chromatography/electron impact ionization mass spectrometry (CGC/EIMS)
chromatogram or the calibration standard solution required for quantitation of PCBs by homolog.
This chromatogram includes PCBs representative of each homolog, three carbon-13 labeled surrogates,
and the deuterated internal standard. The concentration of a]1 components and the CGC/EIMS
parameters are presented in Tables 3, 4, 5, and 7.
-------
TABLE 11. CHARACTERISTIC SIM IONS FOR PCBs
Ion (relative intensity)
Homolog .. Primary Secondary Tertiary
C12H9C1 188 (100)
C12H8C12 222 (100)
Ci2H7Cl3 256 (100)
C12H6C14 292 (100)
Ci2H5Cl5 326 (100)
Ci2H4Cl6 360 (100)
C12H3C17 394 (100)
C12H2C18 430 (100)
Ci2HCl9 464 (100)
C12Clio 498 (100)
190 (33)
224 (66)
258 (99)
290 (76)
328 (66)
362 (82)
396 (98)
432 (66)
466 (76)
500 (87)
-
226 (11)
260 (33)
294 (49)
324 (61)
364 (36)
398 (54)
428 (87)
462. (76)
496 (68)
Source: Rote, J. W., and W. J. Morris, "Use of Isotopic Abundance Ratios in
Identification of Polychlorinated Biphenyls by Mass Spectrometry,"
J. Assoc. Offic. Anal. Chero.. 56(1), 188-199 (1973).
B-31
-------
TABLE 12. LIMITED MASS SCANNING (LMS) RANGES FOR PCBs
Compound
v< 1 ^ITQ \j J_ 1
LJ 1 Oil O \j 1 O
C12H7C13
C12H6C13
Ci2H5Cl5
C12H4C16
C12H3C17
C12H2C18
C12HC19
C12C110
C12D6C14
13C612C6H9C1
13C12H6C14
13C12H2C18
13C12Clio
Mass range (m/z)
186-190
220-226
254-260
288-294
322-328
356-364
386-400
426-434
460-468
494-504
294-300
192-196
300-306
438-446
506-516
a Adapted from Tindall, G. W., and P. E. Wininger, "Gas Chromatography-Mass
Spectrometry Method for Identifying and Determining Polychlorinated Bi-
phenyls," J. Chromatogr., 196, 109-119 (1980).
B-32
-------
11.1.2 If one or the other of these criteria is not met, inter-
. - ferences may have affected the results, and a reanalysis
using full scan EIMS conditions is recommended.
11.2 Full scan data
11.2.1 The peak must elute within the retention time windows
set for that homolog (as described in Section 7.5).
11.2.2 The unknown spectrum must match that of an authentic PCB.
The intensity of the three largest ions in the molecular
cluster (two largest for monochlorobiphenyls) must match
the natural ratio within ±20%. Fragment clusters with
proper intensity ratios must also be present.
11.2.3 Alternatively, a spectral search may be used to auto-
matically reduce the data. The criteria for acceptable
identification include a high index of similarity. For
the -Incos 2300, a fit of 750 or greater must be obtained.
11.3 Disputes in interpretation - Where there is reasonable doubt as
to the identity of a peak as a PCB, the analyst must either iden-
tify the peak as a PCB or proceed to a confirmational analysis
(see Section 13.0).
12.0 Quantitative Data Reduction
12.1 Once a chromatographic peak has been identified as a PCB, the com-
pound is quantitated based either on the integrated abundance of
the SIM data or EICP for the primary characteristic ion in Tables
11 and 12. If interferences are observed for the primary ion,
use the secondary and then tertiary ion for quantitation. If
interferences in the parent cluster prevent quantitation, an ion
from a fragment cluster (e.g., M-70) may be used. Whichever ion
is used, the RF must be determined using that ion. The same cri-
teria should be applied to the surrogate compounds (Table 13).
12.2 Using the appropriate analyte-internal standard pair and response
factor (RF ) as determined in Section 7.3, calculate the concen-
tration ofpeach peak using Equation 12-1.
A) i M V
Concentration (fjg/g) = -E_ . ^- . gi£ . ^£ Eq. 12-1
is p e i
where A = area of the characteristic ion for the analyte PCB
P peak
A. = area of the characteristic ion for the internal
standard peak
RF = response factor of a given PCB congener
B-33
-------
TABLE 13. CHARACTERISTIC IONS FOR 13C-LABELED PCB SURROGATES
" Ion (relative intensity)
Specific compound Primary Secondary Tertiary
13C612C6H9C1 194 (100) 196 (33)
13C12H6C14 304 (100) 306 (49) 302 (78)
13C12H2C18 442 (100) 444 (65) 440 (89)
13C12C110 510 (100) 512 (87) 514 (50)
B-34
-------
M. = mass of internal standard injected (micrograms)
XS ( , •
M = mass of sample extracted (grams)
V. = volume injected (microliters)
V = volume of sample extract (microliters)
12.3 If a peak appears to contain non-PCB interferences, which cannot
be circumvented by a secondary or tertiary ion, either:
12.3.1 Reanalyze the sample on a different column which sepa-
rates the PCB and interf erents ;
12.3.2 Perform additional chemical cleanup (Section 9) and then
reanalyze the sample; or
12.3.3 Quantitate the entire peak as PCB.
12.4 Calculate the recovery of the four 13C surrogates using the ap-
propriate surrogate-internal standard pair and response factor
(RF. ) as determined in Section 7.4 using Equation 12-2.
xs
A M.
Recovery (%)=-£_. ^- . -*i . 100 Eq. 12-2
IS S S
where A = area of the characteristic ion for the surrogate peak
S
A. = area of the characteristic ion for the internal standard
is ,
peak
RF = response factor for the surrogate compound with respect
to the internal standard (Equation 7-2)
M. = mass of internal standard injected (nanograms)
XS
M = mass of surrogate, assuming 100% recovery (nanograms)
S
12.5 Correct the concentration of each peak using Equation 12-3. This
is the final reportable concentration.
Corrected concentration (pg/g) = * 10° ^ 12'3
12.6 Sum all of the peaks for each homolog, and then sum those to yield
the total PCB concentration in the sample. Report all numbers in
Mg/g. The reporting form in Table 14 may be used. If an alter-
nate reporting format (e.g., concentration per peak) is desired,
a different report form may be used. The uncorrected concentra-
tions, percent recovery, and corrected recovery are to be reported.
12.7 Round off all numbers reported to two significant figures.
B-35
-------
TABLE 14. ANALYSIS REPORT
INCIDENTAL PCBs IN COMMERCIAL PRODUCTS OR PRODUCT WASTES
Sample No.
Sample Matrix
Sample Source
Notebook No. or File Location
Volume Extracted
Extraction/Cleanup Procedure
Int. Std.
4-Cl(d6)
Mass Added (|Jg)
(Circle one)
298 300
Ratio
100/49
Intensity
Surrogates
1-C1
4-C1
8-C1
10-C1
Mass Added (|Jg)
(Circle one)
194 196
304 306
442 444
510 512
Ratio
100/33
100/49
100/65
100/87
Intensity % Recovery
(continued)
B-36
-------
TABLE 14 (continued)
Analyte 1° 2° Il°
1-C1 188 190
2-C1 222 224
3-C1 256 258
4-C1 292 290
5-C1 326 328
6-C1 360 362
7-C1 394 396
8-C1 430 432
9-C1 464 466
10-C1 498 500
Total
Reported by:
Name
Signature/Date
Organization
Qualitative
2° Ratio Theoretical
100/33
100/66
100/99
100/76
100/66
100/82
100/98
100/66
100/76
100/87
Internal Audit:
Name
Signature/Date
Organization
Quantitative
Uncorr. Corr.
Ion Cone. Cone.
OK? Used RF (jjg/g) (Mg/g)
Mg/g Mg/g
Uncorr. Corr.
EPA Audit:
Name
Signature/Date
Organization
B-37
-------
13.0 Confirmation
If there is reason to question the qualitative identification (Section
11.0), the analyst may choose to confirm that a peak is not a PCB. Any
technique may be chosen provided that it is validated as having equiva-
lent or superior selectivity and sensitivity to GC/EIMS. Some candidate
techniques include alternate GC columns (with EIMS detection), GC/CIMS,
GC/NCIMS, high resolution EIMS, and MS/MS techniques. Each laboratory
must validate confirmation techniques to show equivalent or superior
selectivity between PCBs and interferences and sensitivity (limit of
quantitation, LOQ).
If a peak is confirmed as being a non-PCB, it may be deleted from the
calculation (Section 12). If a peak is confirmed as containing both
PCB and non-PCB components, it must be quantitated according to Section
12.3.
14.0 Quality Control
14.1 Each laboratory that uses this method must operate a formal qual-
ity control (QC) program. The minimum requirements of this pro-
gram consist of an initial demonstration of laboratory capability
and the analysis of spiked samples as a continuing check on perfor-
mance. The laboratory must maintain performance records to define
the quality of data that are generated. After a date specified by
the Agency, ongoing performance checks should be compared with
established performance criteria to determine if the results of
analyses are within accuracy and precision limits expected of the
method.
14.2 The analysts must certify that the precision and accuracy of the
analytical results are acceptable by:
(
14.2.1 The absolute precision of surrogate recovery, measured
as the RSD of the integrated EIMS area (A ) for a set
of samples, must be ±10%.
14.2.2 The mean recovery (R ) of at least four replicates of a
QC check sample to be supplied by the Agency must meet
Agency-specified accuracy and precision criteria. This
forms the initial data base for establishing control
limits (see Section 14.3 below).
14.3 Control limits - The laboratory must establish control limits
using the following equations:
Upper control limit (UCL) = R +3 RSD
Upper warning limit (UWL) = R +2 RSD
Lower warning limit (LWL) = R - 2 RSD
Lower control limit (LCL) = RC - 3 RSDc
B-38
-------
These may be plotted on control charts. If an analysis of a check
sample falls outside the warning limits, the analyst should be
alerted that potential problems may need correction. If the re-
sults for a check sample fall outside the control limits, the lab-
oratory must take corrective action and recertify the performance
(Section 14.2) before proceeding with analyses. The warning and
control limits should be continuously updated as more check sample
replicates are added to the data base.
14.4 Before processing any samples, the analyst should demonstrate
through the analysis of a reagent blank that all glassware and
reagent interferences are under control. Each time a set of sam-
ples is analyzed or there is a change in reagents, a laboratory
reagent blank should be processed as a safeguard against contam-
ination.
14.5 Procedural QC - The various steps of the analytical procedure
should have quality control measures. These include but are not
limited to:
14.5.1 GC performance - See Section 7.1 for performance criteria.
14.5.2 MS performance - See Section 7.2 for performance criteria.
14.5.3 Qualitative identification - At least 10% of the PCB
identifications, as well as any questionable results,
should be confirmed by a second mass spectrometrist.
14.5.4 Quantitation - At least 10% of all manual calculations,
including peak area calculations, must be checked. After
changes in computer quantitation routines, the results
should be manually checked.
14.6 A minimum of 10% of all samples, one sample per month or one sam-
ple per matrix type, whichever is greater, selected at random, must
be run in triplicate to monitor the precision of the analysis. An
RSD of ±30% or less must be achieved. If the precision is greater
than ±30%, the analyst must be recertified (see Section 14.2).
14.7 A minimum of 10% of all samples, one sample per month or one sam-
ple per matrix type, whichever is greater, selected at random, must
be analyzed by the standard addition technique. Two aliquots of
the sample are analyzed, one "as is" and one spiked (surrogate
spiking and equilibration techniques are described in Section 9.2)
with a sufficient amount of Solution CSxxx to yield approximately
100 pg/g of each compound. The samples are analyzed together and
the quantitative results calculated. The recovery of the spiked
compounds (calculated by difference) must be 80-120%. If the sam-
ple is known to contain specific PCB isomers, these isomers may be
substituted for solution CSxxx. If the concentrations of PCBs are
known to be high or low, the amount added should be adjusted so
that the spiking level is 1.5 to 4 times the measured PCB level
in the unspiked sample.
B-39
-------
14.8 Interlaboratory comparison - Interlaboratory comparison studies
are planned. Participation requirements, level of performance,
and the identity of the coordinating laboratory will be presented
in later revisions.
14.9 It is recommended that the participating laboratory adopt addi-
tional QC practices for use with this method. The specific prac-
tices that are most productive depend upon the needs of the lab-
oratory and the nature of the samples. Field duplicates or trip-
licates may be analyzed to monitor the precision of the sampling
technique. Whenever possible, the laboratory should perform
analysis of standard reference materials and participate in rele-
vant performance evaluation studies.
15.0 Quality Assurance
Each participating laboratory must develop a quality assurance plan ac-
cording to EPA guidelines.5 The quality assurance plan must be submitted
to the Agency for approval.
16.0 Method Performance
The method performance is being evaluated. Limits of quantitation;
average intralaboratory recoveries, precision, and accuracy; and inter-
laboratory recoveries, precision, and accuracy will be presented.
17.0 Documentation and Records
Each laboratory is responsible for maintaining full records of the analy-
sis. Laboratory notebooks should be used for handwritten records. GC/MS
data must be archived on magnetic tape, disk, or a similar device. Hard
copy printouts may be kept in addition if desired. QC records should
be maintained separately from sample analysis records.
The documentation must describe completely how the analysis was performed.
Any variances from the protocol must be noted and fully described. Where
the protocol lists options (e.g., sample cleanup), the option used and
specifics (solvent volumes, digestion times, etc.) must be stated.
B-40
-------
REFERENCES
1. "Methods 330.4 (Titrimetric, DPD-FAS) and 330.5 (Spectrophotometric, DPD)
for Chlorine, Total Residual," Methods for Chemical Analysis of Water and
Wastes, U.S. Environmental Protection Agency, Environmental Monitoring
and Support Laboratory, Cincinnati, Ohio, March 1979, EPA 600-4/79-020.
2. Erickson, M. D., and J. S. Stanley, "Methods of Analysis for Incidentally
Generated PCBs—Literature Review and Preliminary Recommendations," Interim
Report No. 1, EPA Contract No. 68-01-5915, Task 51, 1982.
3. Bellar, T. A., and J. J. Lichtenberg, "The Determination of Polychlorinated
Biphenyls in Transformer Fluid and Waste Oils," Prepared for U.S. Environ-
mental Protection Agency, (1981) EPA-600/4-81-045.
4. American Society for Testing and Materials, "Standard Method for Analysis
of Environmental Materials for Polychlorinated Biphenyls," pp. 877-885 in
Annual Book of ASTM Standards, Part 40, Philadelphia, Pennsylvania (1980).
ANSI/ASTM D 3304 - 77.
5. "Quality Assurance Program Plan for the Office of Toxic Substances,"
Office of Pesticides and Toxic Substances, U.S. Environmental Protection
Agency, Washington, B.C., October 1980.
B-41
-------
APPENDIX C
ANALYTICAL METHOD: THE ANALYSIS OF BY-PRODUCTS
CHLORINATED BIPHENYLS IN AIR
C-l
-------
THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS IN AIR
1.0 Scope and Application
1.1 This is a gas chromatographic/electron impact mass spectrometric
(GC/EIMS) method applicable to the determination of chlorinated
biphenyls (PCBs) in air emitted from commercial production through
stacks, as fugitive emissions, or static (room, other containers,
or outside) air. The PCBs present may originate either as syn-
thetic by-products or as contaminants derived from commercial PCB
products (e.g., Aroclors). The PCBs may be present as single
isomers or complex mixtures and may include all 209 congeners
from monochlorobiphenyl through decachlorobiphenyl listed in
Table 1.
1.2 The detection and quantitation limits are dependent upon the vol-
ume of sample collected, the complexity of the sample matrix and
the ability of the analyst to remove interferents and properly
maintain the analytical system. The method accuracy and preci-
sion will be determined in future studies.
1.3 This method is restricted to use by or under the supervision of
analysts experienced in the use of gas chromatography/mass spec-
trometry (GC/MS) and in the interpretation of gas chromatograms
and mass spectra. Prior to sample analysis, each analyst must
demonstrate the ability to generate acceptable results with this
method by following the procedures described in Section 14.2.
1.4 The validity of the results depends on equivalent recovery of the
analyte and 13C PCBs. If the 13C PCBs are not thoroughly incor-
porated in the matrix, the method is not applicable.
1.5 During the development and testing of this method, certain an-
alytical parameters and equipment designs were found to affect
the validity of the analytical results. Proper use of the method
requires that such parameters or designs must be used as speci-
fied. These items are identified in the text by the word "must."
Anyone wishing to deviate from the method in areas so identified
must demonstrate that the deviation does not affect the validity
of the data. Alternative test procedure approval must be ob-
tained from the Agency. An experienced analyst may make modifi-
cations to parameters or equipment identified by the term "recom-
mended." Each time such modifications are made to the method,
the analyst must repeat the procedure in Section 14.2. In this
case, formal approval is not required, but the documented data
from Sectin 14.2 must be on file as part of the overall quality
assurance program.
C-2
-------
TABLE 1. NUMBERING OF PCB CONGENERS3
NO.
1
3
4
5
6
7
a
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
23
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Structure
ManochloroblphtriYU
2
4
D1cMoroMpti«ny1>
2.2'
2,3
2.3'
2.4
2.4'
.5
,6
,3'
,4
,4'
.5
•».«'
TrtchlorolilphenyU
2.2', 3
2.2'. 4
2.2'. 5
2.2'. 6
2,3,3'
2.3,4
2.3.4'
2.3.5
2.3,6
2,3', 4
2.3'. 5
2, 3', 6
2,4,4'
2.4.5
2,4.6
2. 4'. 5
2,4', 6
2'. 3. 4
2'. 3,5
3.3',4
3.3',5
3.4,4'
3.4.5
3. 4'. 5
Tttracft 1 oroftl oh«nyl s
2,2' ,3,3'
2,2'. 3.4
2.2', 3,4'
2.2'.3,5
2,2',3,5'
2,2', 3, 6
2.2'. 3,6'
2.2', 4,4'
2.2', 4,5
2.2', 4.5'
2.2', 4,6
2,2', 4.6'
NO.
52
S4
55
56
57
58
59
60
61
62
63
M
65
66
67
68
69
70
71
72
73
7«
75
76
77
78
79
80
31
82
83
84
85
86
87
88
99
90
91
92
93
9*
95
96
97
98
99
100
101
102
103
104
structurt
T«tricftlofnb1ph»nylt
2.2'. 5,5'
29* e ic*
• & |9*O
2.2'. 6,6'
2,3. 3'N
2.3,3',4'
2.3,3' .S
2.3.3'. 5'
2.3,3', 6
2.3,4,4'
2,3,4.5
2,3.4,6
2.3,4'.S
2. 3,4', 6
2.3,5.6
2.3', 4,4'
2, 3', 4,5
2.3', 4,5'
2.3'. 4,6
2,3'.V.S
2,3',4>,6
2.3' ,5.5'
2. 3', 5', 6
2.4, 4'. S
2.4.41.6
2'. 3 4,$
3.3' 4.4'
3.3' 4,5
3.3' 4.5'
3.3' 5,5'
3.4. 41, 5
P«ntidi1oro61phenyl«
2,2',3,3',4
2. 2', 3, 3', 5
2.2' .3, 3' ,6
2.2' .3,4,4'
2.2',3,4.5
2,2'.3,4.5'
2.2', 3,4,6
2.2'. 3,4, 6'
2.2'.3.4'.5
2, 2'. 3,4'. 6
2.2' ,3.5. 5'
2,2'. 3, 5, 6 •>
2.2'. 3,5,6'
2,2'. 3.5', 8
2.2',3.6.6'
2.2' ,3' .4,5
2,2' ,3'. 4,6
2,2',4,4'15
2,2'. 4. 4' .6
2.2', 4,5.5'
2,2'.4.5,6'
2,2'. 4,5'. 6
2,2' .4.6.6'
No.
105
1/VE
IUD
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
Structure
Pentich 1 orob t ohefiy 1 s
2.3,3'. 4,4'
24 ^, 1 t
,3,3 .4,9
2, 3,3' .4'. 5
2,3,3', 4. 5'
2.3, 3', 4, 6
2. 3, 3' ,4', 6
2.3,3' ,5,5'
2, 3, 3', 5,6
2.3.3'. 5'.6
2.3,4. 41 .5
2.3.4,4'.$
2.3.4.5.6
2.3.4' ,5,6
2.3'.4.4'.S
2.3'. 4,4'. 6
2,3', 4,5, 5'
2,3', 4. 5', 6
21. 3,3'. 4.5
2'. 3. 4. 4'. 5
2'. 3. 4,5. 5'
2'.3.4.5.6'
3, 3', 4. 4'. 5
3, 3', 4, 5, 5'
Htmehlorabfphenyli
2, 2', 3, 3'. 4,4'
2.2', 3, 3' .4.5
2.2', 3. 3'. 4, 5'
2.2'.3.3',4.6
2.2', 3.3'. 4,6'
2. 2'. 3, 3'. 5, 5'
2, 2', 3, 3', 5, 6
2.2', 3, 3'. 5, 6'
2, 2', 3, 3'. 6, 6'
2.2' .3,4,4' ,5
2,2'. 3, 4, 4'. 5'
2.2' .3,4, 4' ,6
2,2', 3, 4,4'. 6'
2.2',3.4,5,5'
2.2'. 3, 4, 5,6
2.2',3.4.5.6'
2.2'.3,4.5',6
2.2'.3,4,6.6'
2.2'. 3.4'. 5.5'
2.2'.3,4',5.6
2.2'. 3. 4'. 5. 6'
2.2'. 3,4'. 5' .6
2,2'. 3,4'. 6.6'
2.2'.3,S,5',6
2.2'. 3,5.6,6'
2.2' ,4, 4', 5.5'
2,2'. 4,4', 5.6'
2.2'. 4, 4'. 6,6'
2.3,3'. 4,4' ,5
2.3. 3' .4, 4'. 5'
2,3,3'.4.4',6
2.3,3' .4.5. 5'
2.3,3'. 4,5,6
NO.
161
1X9
IO*
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
structure
Hemch! orobl phenyl i
2. 3. 3'. 4. 5'. 6
2.3,3' ,4'. 5,6
2, 3, 3'. 4'. 5' .6
2.3.3', 5, 5'. 6
2, 3, 4, 4'. 5, 6
2.3'. 4, 4'. 5. 5'
2. 3' ,4, 4'. 5'. 6
3,3' .4. 4'. 5,5'
Hetmchlorobtphenyl t
2.2' .3. 3'. 4. 4', 5
2, 2'. 3, 3' ,4. 41. 6
2. 2'. 3. 3'. 4. 5, 5'
2, 2', 3, 3', 4, 5, 6
2, 2'. 3. 3'. 4, 5, 6'
2.2'.3,3'.4,51,6
2,2'.3.3'.4.6.6'
2, 2'. 3, 3'. 4'. 5, 6
2, 2', 3. 3', 5, 5'. 6
2, 2', 3.4, 4'. 5. 6'
2. 2'. 3,4,4'. 5'. 5
2. 2'. 3, 4, 4'. 6, 6'
2,2' ,3, 4.5, 5'. 6
2, 2'. 3, 4, 5. 6, 6'
2, 2', 3, 4', 5. 5'. 6
2. 2', 3. 41. 5, 6, 6'
2. 3, 3', 4. 4'. 5, 5'
2, 3, 3'. 4, 4', 5. 6
2, 3, 3', 4, 4' ,5', 5
2, 3, 3', 4, 5,5', 6
2.3,3',4'.5,51,S
OctiCBlorobipftenyl i
2,2' .3. 3' .4, 4', 5, 5'
2,2', 3, 3', 4, 4'. 5,0
2,2' .3, 3', 4, 4'. 3,6'
2. 2'. 3, 3', 4, 4', 6, 6'
2.2'. 3, 3' ,4, 5, 5' ,o
2. 2', 3. 3'. 4, 5, 6, 6'
2,2',3,3'.4.5',6.6'
2.2'. 3, 3', 4. 5, 5'. 6'
2.2* .3.3' .5,5', 6,6'
2,2' .3,4,4'. 5,5',6
2.2'. 3,4, 4'. 5, 6, 6'
2.3.3'.4,4'.5,5',6
Honieh 1 orob< oneny 1 ?
2.2'.3,3'.4,4',S.5'.6
2, 2' ,3, 3' .4, 4', 5, 6,6'
2,2'. 3. 3'. 4, 5, 5'. 6. 6'
Oecacft 1 orom oheny 1
Z^'.J.JU.i'.s.s'.e.s1
•Adopttd fro» BtllscMwr. X. ind 2*11, M., Fr««n1ui Z. Anal.Omi., 302. 20-31 (1980).
C-3
-------
2.0 Summary
2.1 The air must be sampled such that the specimen collected for
analysis is representative of the whole. Statistically designed
selection of the sampling position (stack, flue, port, etc.) or
time should be employed. Gaseous and particulate PCBs are with-
drawn isokinetically from stacks, room air exhausts, process point
exhausts, and other flowing gaseous streams using a sampling train.1
The PCBs are collected in the Florisil adsorbent tube and in the
impingers in front of the adsorbent. PCBs are sampled from ambient
air and other static gaseous sources onto a Florisil adsorbent
tube. The sample must be preserved to prevent PCB loss prior to
analysis. Storage at 4°C is recommended.
2.2 The Florisil adsorbent is extracted with hexane in a Soxhlet ex-
tractor, the aqueous condensate is extracted with hexane and the
acetone/hexane impinger rinse is back-extracted with water. All
three organic extracts are then combined. Optional cleanup tech-
niques may include sulfuric acid cleanup and Florisil adsorption
chromatography. The sample is concentrated to a final known vol-
ume for instrumental determination.
2.3 The PCB content of the sample extract is determined by capillary
(preferred) or packed column gas chromatography/electron impact
mass spectrometry (CGC/EIMS or PGC/EIMS) operated in the selected
ion monitoring (SIM), full scan, or limited mass scan (LMS) mode.
2.4 PCBs are identified by comparison of their retention time and mass
spectral intensity ratios to those in calibration standards.
2.5 PCBs are quantitated against the response factors for a mixture
of 11 PCB congeners using the internal standard technique.
2.6 The PCBs identified by the SIM technique may be confirmed by full
scan CGC/EIMS, retention on alternate GC columns, other mass spec-
trometric techniques, infrared spectrometry, or other techniques,
provided that the sensitivity and selectivity of the technique are
demonstrated to be comparable or superior to GC/EIMS.
2.7 The analysis time is dependent on the extent of workup employed.
The time required for instrumental analysis of a single sample
excluding data reduction and reporting, is about 30 to 45 min.
2.8 Appropriate quality control (QC) procedures are included to assess
the performance of the analyst and estimate the quality of the re-
sults . These QC procedures include the demonstration of laboratory
capability: periodic analyst certification, the use of control
charts, and the analysis of blanks, replicates, and standard addi-
tion samples. A quality assurance (QA) plan must be developed for
each laboratory.
C-4
-------
2.9 While several options are available throughout this method, the
recommended procedure for stack gases to be followed is:
2.9.1 The sample is collected using a modified Method 5 train1
according to a scheme which permits extrapolation of the
sample data to the source being assessed.
2.9.2 The sample is preserved at 4°C to prevent any loss of
PCBs or changes in matrix which may adversely affect re-
covery.
2.9.3 The three sample fractions are extracted and combined.
2.9.4 The extract is cleaned up and concentrated to an appro-
priate volume. Internal standards are added.
2^9.5 An aliquot of the extract is analyzed by CGC/EIMS oper-
ated in the SIM mode. On-column injections onto a 15-m
DB-5 capillary column, programmed (for toluene solutions)
from 110° to 325°C at 10°/min after a 2 min hold is used.
Helium at 45-cm/sec linear velocity is used as the car-
rier gas.
2.9.6 PCBs are identified by retention time and mass spectral
intensities.
2.9.7 PCBs are quantitated against the response factors for a
mixture of 11 PCB congeners.
2.9.8 The total PCBs are obtained by summing the amounts for
each homolog found, and the concentration is reported
as micrograms per cubic meter.
3.0 Interferences
3.1 Method interferences may be caused by contaminants, in sample col-
lection media, solvents, reagents, glassware, and other sample
processing hardware, leading to discrete artifacts and/or ele-
vated baselines in the total ion current profiles. All of these
materials must be routinely demonstrated to be free from interfer-
ences by the analysis of laboratory reagent blanks as described
in Section 14.4.
3.1.1 Glassware must be scrupulously cleaned. All glassware
is cleaned as soon as possible after use by rinsing with
the last solvent used. This should be followed by deter-
gent washing with hot water and rinses with tap water and
reagent water. The glassware should then be drained dry
and heated in a muffle furnace at 400°C for 15 to 30 min.
Some thermally stable materials, such as PCBs, may not
be eliminated by this treatment. Solvent rinses with
acetone and pesticide quality hexane may be substituted
C-5
-------
for the muffle furnace heating. Volumetric ware should
not be heated in a muffle furnace. After it is dry and
cool, glassware should be sealed and stored in a clean
environment to prevent any accumulation of dust or other
contaminants. It is stored inverted or capped with
aluminum foil.
3.1.2 The use of high purity reagents and solvents helps to
minimize interference problems. Purification of solvents
by distillation in all-glass systems may be required.
All solvent lots must be checked for purity prior to use.
3.2 Matrix interferences may be caused by contaminants that are coex-
tracted from the sorbent material or impingers. The extent of
matrix interferences will vary considerably from source to source,
depending upon the nature and diversity of the sources of samples.
4.0 Safety
4.1 The toxicity or carcinogenicity of each reagent used in this
method has not been precisely defined; however, each chemical
compound should be treated as a potential health hazard. From
this viewpoint, exposure to these chemicals must be reduced to
the lowest possible level by whatever means available. The lab-
oratory is responsible for maintaining a current awareness file
of OSHA regulations regarding the safe handling of the chemical
specified in this method. A reference file of material data han-
dling sheets should also be made available to all personnel in-
volved in the chemical analysis.
4.2 Polychlorinated biphenyls have been tentatively classified as
known or suspected human or mammalian carcinogens. Primary
standards of these toxic compounds should be prepared in a hood.
Personnel must wear protective equipment, including gloves and
safety glasses.
Congeners highly substituted at the meta and para positions and
unsubstituted at the ortho positions are reported to be the most
toxic. Extrme caution should be taken when handling these com-
pounds neat or in concentrated solution. The class includes
3,3',4'4'-tetrachlorobiphenyl (both natural abundance and isotop-
ically labeled).
4.3 Waste disposal must be in accordance with RCRA and applicable
state rules.
5.0 Apparatus and Materials
All specifications are suggestions only. Catalog numbers and suppliers
are included for illustration only.
C-6
-------
5.1 Stack sampling train1 - See Figure 1; a series of four impingers
with a solid adsorbent trap between the third and fourth impingers.
The train may be constructed by adaptation from a Method 5 train.2
Descriptions of the train components are contained in the follow-
ing subsections.
5.1.1 Probe nozzle - Stainless steel (316) with sharp, tapered
leading edge. The angle of taper shall be ^ 30° and the
taper shall be on the outside to preserve a constant in-
ternal diameter. The probe nozzle shall be of the button-
hook or elbow design, unless otherwise specified by the
Agency. The wall thickness of the nozzle shall be less
than or equal to that of 20 gauge tubing, i.e., 0.165 cm
(0.065 in.) and the distance from the tip of the nozzle
to the first bend or point of disturbance shall be at
least two times the outside nozzle tubing. Other con-
figurations and construction material .may be used with
approval from the Agency.
5.1.2 Probe liner - Borosilicate or quartz glass equipped with
a connecting fitting that is capable of forming a leak-
free, vacuum tight connection without sealing greases;
such as Kontes Glass Company "0" ring spherical ground
ball joints (model K-671300) or University Research
Glassware SVL teflon screw fittings.
A stainless steel (316) or water-cooled probe may be used
for sampling high temperature gases with approval from
the Agency. A probe heating system may be used to prevent
moisture condensation in the probe.
5.1.3 Pitot tube - Type S, or equivalent, attached to probe to
allow constant monitoring of the stack gas velocity.
The face openings of the pitot tube and the probe nozzle
shall be adjacent and parallel to each other but not
necessarily on the same plane, during sampling. The free
space between the nozzle and pitot tube shall be at least
1.9 cm (0.75 in.). The free space shall be set based on
a 1.3 cm (0.5 in.) ID nozzle, which is the largest size
nozzle used.
The pitot tube must also meet the criteria specified in
Method 22 and be calibrated according to the procedure
in the calibration section of that method.
5.1.4 Differential pressure gauge - Inclined manometer capable
of measuring velocity head to within 10% of the minimum
measured value. Below a differential pressure of 1.3 mm
(0.05 in.) water gauge, micromanometers with sensitivities
of 0.013 mm (0.0005 in.) should be used. However, micro-
manometers are not easily adaptable to field conditions
and are not easy to use with pulsating flow. Thus, other
methods or devices acceptable to the Agency may be used
when conditions warrant.
C-7
-------
Stack
Wall
Thermometer'
Florisil Tube
Check
Valve
Probe (r^
£
Reverse-Type
Pitor Tube
Control Box
Figure 1. PCB sampling train for stack gases.
C-J
-------
5.1.5 Impingers - Four impingers with connecting fittings able
to form leak-free, vacuum tight seals without sealant
greases when connected together as shown in Figure 1.
The first and second impingers are of the Greenburg-
Smith design. The final two impingers are of the
Greenburg-Smith design modified by replacing the tip
with a 1.3 cm (1/2 in.) ID glass tube extending to 1.3
cm (1/2 in.) from the bottom of the flask.
One or two additional modified Greenburg-Smith impingers
may be added to the train between the third impinger and
the Florisil tube to accommodate additional water col-
lection when sampling high moisture gases. Throughout
the preparation, operation, and sample recovery from the
train, these additional impingers should be treated
exactly like the third impinger.
5.1.6 Solid adsorbent tube - Glass with connecting fittings
able to form leak-free, vacuum tight seals without seal-
ant greases (Figure 2). Exclusive of connectors, the
tube has a 2.2 cm inner diameter, is at least 10 cm long,
and has four deep indentations on the inlet end to aid
in retaining the adsorbent. Ground glass caps (or
equivalent) must be provided to seal the adsorbent-filled
tube both prior to and following sampling.
5.1.7 Metering system - Vacuum gauge, leak-free pump, thermom-
eters capable of measuring temperature to within ±3°C
(•^ 5°F), dry gas meter with 2% accuracy at the required
sampling rate, and related equipment, or equivalent, as
required to maintain an isokinetic sampling rate and to
determine sample volume. When the metering system is
used in conjunction with a pitot tube, the system shall
enable checks of isokinetic rates.
i
5.1.8 Barometer - Mercury, aneroid, or other barometers cap-
able of measuring atmospheric pressure to within 2.5 mm
Hg (0.1 in. Hg). In many cases, the barometric reading
may be obtained from a nearby weather bureau station, in
which case the station value shall be requested and an
adjustment for elevation differences shall be applied at
a rate of -2.5 mm Hg (0.1 in. Hg) per 30 mm (100 ft) ele-
vation increase.
5.2 Static air sampling train1 - The sampling train, see Figure 3,
consists of a glass-lined probe, an adsorbent tube containing
Florisil, and the appropriate valving and flow meter controls for
isokinetic sampling as described in Section 5.1. The sampling
apparatus in Figure 3 is the same as that in Figure 1 and Section
5.1, except that the Smith-Greenburg impingers and heated probe
are not used. If condensation of significant quantities of mois-
ture prior to the solid adsorbent is expected, Section 5.1 of the
C-9
-------
J28/12
10cm
j 28/12
Figure 2. Florisil adsorbent tube.
C-10
-------
Probe (to sample from duct) ^—
Glass- lined Probe
a
Type S
Pitot Tube
Manometer
Thermometers
0(7
By - pass
Valve
Florisil
Glass Wool
Check Valve
Vacuum
Gauge
I Integrated |
Flow Meter
Manometer -
Air
Tight
Pump
Vacuum
Line
Figure 3. PCB sampling train for static air.
C-ll
-------
method should be used. Since probes and adsorbent tubes are not
cleaned up in the field, a sufficient number must be provided for
sampling and allowance for breakage.
5.3 Sample recovery
5.3.1 Ground glass caps - To cap off adsorbent tube and the
other sample exposed portions of the train.
5.3.2 Teflon FEP® wash bottle - Two, 500 ml, Nalgene No.
0023A59 or equivalent.
5.3.3 Sample storage containers - Glass bottles, 1 liter, with
TFE®-lined screw caps.
5.3-4 Balance - Triple beam, Ohaus Model 7505 or equivalent.
5.3.5 Aluminum foil - Heavy duty.
5.3.6 Metal can - To recover used silica gel.
5.4 Analysis
5.4.1 Glass Soxhlet extractors - 40 mm ID complete with 45/50
S condenser, 24/40 $ 250 ml round-bottom flask, heating
mantle for 250 ml flask, and power transformer.
5.4.2 Teflon FEP wash bottle - Two, 500 ml, Nalgene No. 0023A59
or equivalent.
5.4.3 Separatory funnel - 1,000 ml with TFE® stopcock.
5.4.4 Kuderna-Danish concentrators - 500 ml.
5.4.5 Steam bath.
5.4.6 Separatory funnel - 50 ml with TFE® stopcock.
5.4.7 Volumetric flask - 25.0 ml, glass.
5.4.8 Volumetric flask - 5.0 ml, glass.
5.4.9 Culture tubes - 13 x 100 mm, glass with TFE®-lined screw
caps.
5.4.10 Pipette - 5.0 ml glass.
5.4.11 Teflon®-glass syringe - 1 ml, Hamilton 1001 TLL or
equivalent with Teflon® needle.
5.4.12 Syringe - 10 (Jl, Hamilton 701N or equivalent.
C-12
-------
5.4.13 Disposable glass pipettes with bulbs - To aid transfer
of the extracts.
5.4.14 Gas chromatography/mass spectrometer system.
5.4.14.1 Gas chromatograph - An analytical system com-
plete with a temperature programmable gas chro-
matograph and all required accessories includ-
ing syringes, analytical columns, and gases.
The injection port must be designed for on-
coluran injection when using capillary columns
or packed columns. Other capillary injection
techniques (split, splitless, "Grob," etc.)
may be used provided the performance specifi-
cations stated in Section 7.1 are met.
5.4.14.2 Capillary GC column - A 12-20 m long x 0.25 mm
ID fused silica column with a 0.25 (J™ thick
DB-5 bonded silicone liquid phase (J&W Scien-
tific) is recommended. Alternate liquid phases
may include OV-101, SP-2100, Apiezon L, Dexsil
300, or other liquid phases which meet the per-
formance specifications stated in Section 7.1.
5.4.14.3 Packed GC column - A 180 cm x 0.2 cm ID glass
column packed with 3% SP-2250 on 100/120 mesh
Supelcoport or equivalent is recommended.
Other liquid phases which meet the performance
specifications stated in Section 7.1 may be
substituted.
5.4.14.4 Mass spectrometer - Must be capable of scanning
from 150 to 550 daltons every 1.5 sec or less,
collecting at least five spectra per chromato-
graphic peak, utilizing a 70-eV (nominal) elec-
tron energy in the electron impact ionizaton
mode and producing a mass spectrum which meets
all the criteria in Table 2 when 50 ng of deca-
fluorotriphenyl phosphine [DFTPP, bis(perfluoro-
phenyl)phenyl phosphine] is injected through
the GC inlet. Any GC-to-MS interface that
gives acceptable calibration points at 10 ng
per injection for each PCB isomer in the cali-
bration standard and achieves all acceptable
performance criteria (Section 10) may be used.
Direct coupling of the fused silica column to
the MS is recommended. Alternatively, GC to
MS interfaces constructed of all glass or glass-
lined materials are recommended. Glass can be
deactivated by silanizing with dichlorodimethyl-
silane.
C-13
-------
TABLE 2. DFTPP KEY IONS AND ION ABUNDANCE CRITERIA
Mass Ion abundance criteria
197 Less than 1% of mass 198
198 100% relative abundance
199 5-9% of mass 198
275 10-30% of mass 198
365 Greater than 1% of mass 198
441 Present, but less than mass 443
442 Greater than 40% of mass 198
443 17-23% of mass 442
C-14
-------
5.4.14.5 A computer system that allows the continuous
acquisition and storage on machine-readable
media of all mass spectra obtained throughout
the duration of the chromatographic program
must be interfaced to the mass spectrometer.
The data system must have the capability of
integrating the abundances of the selected
ions between specified limits and relating
integrated abundances to concentrations using
the calibration procedures described in this
method. The computer must have software that
allows searching any GC/MS data file for ions
of a specific mass and plotting such ion abun-
dances versus time or scan number to yield an
extracted ion current profile (EICP). Software
must also be available that allows integrating
the abundance in any EICP between specified
time or scan number limits.
6.0 Reagents
6.1 Sampling
6.1.1
6.2
6.1.2
6.1.3
6.1.4
Florisil - Floridin Company, 30/60 mesh, Grade A. The
Florisil is cleaned by 8 hr Soxhlet extraction with hex-
ane and then by drying for 8 hr in an oven atvl!0°C and
is activated by heating to 650°C for 2 hr (not to exceed
3 hr) in a muffle furnace. After allowing to cool to
near 110°C transfer the clean, active Florisil to a clean,
hexane-washed glass jar and seal with a TFE®-lined lid.
The Florisil should be stored at 110°C until taken to
the field for use. Florisil that has been stored more
than 1 month must be reactivated before use.
Glass wool - Cleaned by thorough rinsing with hexane,
dried in a 110°C oven, and stored in a hexane-washed
glass jar with TFE®-lined screw cap.
Water - Deionized, then glass-distilled, and stored in
hexane-rinsed glass containers with TFE®-lined screw caps.
Silica gel - Indicating type, 6-16 mesh. If previously
used, dry at 175°C for 2 hr. New silica gel may be used
as received.
6.1.5 Crushed ice.
Solvents - All solvents must be pesticide residue analysis grade.
New lots should be checked for purity by concentrating an aliquot
by at least as much as is used in the procedure.
C-15
-------
6.3 Calibration standard congeners - Standards of the PCB congeners
listed in Table 3 are available from Ultra Scientific, Hope,
Rhode Island; or Analabs, North Haven, Connecticut.
6.4 Calibration standard stock solutions - Primary dilutions of each
of the individual PCBs listed in Table 3 are prepared by weighing
approximately 1-10 mg of material within 1% precision. The PCB
is then dissolved and diluted to 1.0 ml with hexane. The concen-
tration is calculated in mg/ml. The primary dilutions are stored
at 4°C in screw-cap vials with Teflon cap liners. The meniscus
is marked on the vial wall to monitor solvent evaporation. Pri-
mary dilutions are stable indefinitely if the seals are maintained.
The validity of primary and secondary dilutions must be monitored
on a quarterly basis by analyzing four quality control check sam-
ples (see Section 14.2).
6.5 Working calibration standards - Working calibration standards are
prepared that are similar in PCB composition and concentration to
the samples by mixing and diluting the individual standard stock
solutions. Example calibration solutions are shown in Table 3.
The mixture is diluted to volume with pesticide residue analysis
quality hexane. The concentration is calculated in ng/ml as the
individual PCBs. Dilutions are stored at 4°C in narrow-mouth,
screw-cap vials with Teflon cap liners. The meniscus is marked
on the vial wall to monitor solvent evaporation. These secondary
dilutions can be stored indefinitely if the seals are maintained.
These solutions are designated "CSxxx," where the xxx is used to
encode the nominal concentration in ng/ml.
6.6 Alternatively, certified stock solutions similar to those listed
in Table 3 may be available from a supplier, in lieu of the pro-
cedures described in Section 6.4.
6.7 DFTPP standard - A 50 ng/(Jl solution of DFTPP is prepared in ace-
tone or another appropriate solvent.
6.8 Internal standard stock solution - The four 13C-labeled PCBs
listed in Table 4 may be available from a supplier as a certi-
fied solution. This solution may be used as received or diluted
further.
6.9 Solution stability - The calibration standard, surrogate and DFTPP
solutions should be checked frequently for stability. These solu-
tions should be replaced after 6 months, or sooner if comparison
with quality control check samples indicates compound degradation
or concentration change.
6.10 Quality control check samples will be supplied by the Agency.
C-16
-------
TABLE 3. CONCENTRATIONS OF CONGENERS IN PCS CALIBRATION STANDARDS (ng/ml)3
Homo log
1
1
2
3
4
5
6
7
8
9
10
4
1
4
8
10
Congener
no.
1
3
7
30
50
97
143
183
202
207
209
210 (IS)
211 (RS)
212 (RS)
213 (RS)
214 (RS)
CS1000
1,040
1,000
1,040
1,040
1,520
1,740
1,920
2,600
4,640
5,060
4,240
255
104
257
407
502
CS100
104
100
104
104
152
174
192
260
464
506
424
255
104
257
407
502
CS050
52
50
52
52
76
87
96
130
232
253
212
255
104
257
407
502
CS010
10
10
10
10
15
17
19
26
46 .
51
42
255
104
257
407
502
a Concentrations given as examples only.
C-17
-------
TABLE 4. COMPOSITION OF INTERNAL STANDARD SPIKING SOLUTION (SS100)
CONTAINING 13C-LABELED PCBsa
Congener
no.
211
212
213
214
Compound
(I1 ,2' ,3' ,4' ,5' ,6'-13C6)4-chlorobiphenyl
(13Ci2)3>3' ,4,4'-tetrachlorobiphenyl
(13C12)2,2' ,3,3' ,5,5' ,6,6'-octachlorobiphenyl
(13Ci2)decachIorobiphenyl
Concentration
(Mg/ml)
104
257
395
502
a Concentrations given as examples only.
C-18
-------
7.0 Calibration
Maintain a laboratory log of all calibrations.
7.1 Sampling train
7.1.1 Probe nozzle - Using a micrometer, the inside diameter
of the nozzle is measured to the nearest 0.025 mm (0.001
in.). Three separate measurements are made using differ-
ent diameters each time and obtain the average of the
measurements. The difference between the high and low
numbers must not exceed 0.1 mm (0.004 in.).
When nozzles become nicked, dented, or corroded, they
must be reshaped, sharpened, and recalibrated before use.
Each nozzle must be permanently and uniquely identified.
7.1.2 Pitot tube - The pitot tube must be calibrated according
to the procedure outlined in Method 2.2
7.1.3 Dry gas meter and orifice meter - Both meters must be
calibrated according to the procedure outlined in APTD-
0581.3 When diaphragm pumps with bypass valves are used,
proper metering system design is checked by calibrating
the dry gas meter at an additional flow rate of 0.0057
m3/min (0.2 cfm) with the bypass valve fully opened and
then with it fully closed. If there is more than ±2%
difference in flow rates when compared to the fully
closed position of the bypass valve, the system is not
designed properly and must be corrected.
7.1.4 Probe heater calibration - The probe heating system must
be calibrated according to the procedure contained in
APTD-0581.3
7.1.5 Temperature gauges - Dial and liquid filled bulb thermom-
eters are calibrated against mercury-in-glass thermometers,
Thermocouples should be calibrated in constant tempera-
ture baths.
7.2 The gas chromatograph must meet the minimum operating parameters
shown in Tables 5 and 6, daily. If all of the criteria are not
met, the analyst must adjust conditions and repeat the test until
all criteria are met.
7.3 The mass spectrometer must meet the minimum operating parameters
shown in Tables 2, 7, and 8, daily. If all criteria are not met,
the analyst must retune the spectrometer and repeat the test un-
til all conditions are met.
C-19
-------
TABLE 5. OPERATING PARAMETERS FOR CAPILLARY COLUMN GAS CHROMATOGRAPHIC SYSTEM
Parameter
Recommended
Tolerance
Gas chromatograph
Column
Liquid phase
Liquid phase thickness
Carrier gas
Carrier gas velocity
Injector
Injector temperature
Injection volume
Initial column temperature
Column temperature program
Finnigan 9610
15 ra x 0.255 mm ID
Fused silica
DB-5 (J&W)
0.25 pm
Helium
/i- / b
45 cm/sec
On-column (J&W)C
Optimum performance
1.0 Ml°
70°C (2 min)d
70°-325°C at 10°C/min£
Other
Other
Other nonpolar
or semipolar
< 1 pm
Hydrogen
Optimum performance
Other
Optimum performance
Other
Other
Other
Separator
Transfer line temperature
Tailing factorh
Peak width1
None1"
280°C
0.7-1.5
7-10 sec
Glass jet or othe
Optimum**
0.4-3
< 15 sec
a Substitutions permitted with any common apparatus or technique provided
performance criteria are met.
b Measured by injection of air or methane at 270°C oven temperature.
c For on-column injection, manufacturer's instructions should be followed
regarding injection technique. . ,
d With on-column injection, initial temperature equals boiling point of the
solvent; in this instance, hexane.
e Cj2Cli6 elutes at 270°C. Programming above this temperature ensures a
clean column and lower background on subsequent runs.
f Fused silica columns may be routed directly into the ion source to prevent
separator discrimination and losses.
g High enough to elute all PCBs, but not high enough to degrade the column
if routed through the transfer line.
h Tailing factor is width of front half of peak at 10% height divided by
width of back half of peak at 10% height for single PCB congeners in solu-
tion CSxxx.
i Peak width at 10% height for a single PCB congener is CSxxx.
C-20
-------
TABLE 6. OPERATING PARAMETERS FOR PACKED COLUMN GAS CHROMATOGRAPHY SYSTEM
Parameter
Recommended
Tolerance
Gas chromatograph
Column
Finnigan 9610
180 cm x 0.2 cm ID
Other3
Other
Column packing
Carrier gas
Carrier gas flow rate
Injector
Injector temperature
Injection volume
Initial column temperature
Column temperature program
Separator
Transfer line temperature
Tailing factor
Peak widthd
glass
3% SP-2250 on 100/
120 mesh Supelcoport
Helium
30 ml/min
On-column
250°C
1.0 pi
150°C, 4 min
150°-260°C at 8°/min
Glass jet
280°C
,0.7-1.5
10-20 sec
Other nonpolar
or semipolar
Hydrogen
Optimum performance
Other
Optimum
^ 5 Ml
Other
Other
Other
Optimum3
0.4-3
< 30 sec
a Substitutions permitted if performance criteria are met.
b High enough to elute all PCBs.
c Tailing factor is width of front half of peak at 10% height divided by
width of back half of peak at 10% height for single PCB congeners in solu-
tion CSxxx.
d Peak width at 10% height for a single PCB congener is CSxxx.
C-21
-------
TABLE 7. OPERATING PARAMETERS FOR QUADRUPOLE MASS SPECTROMETER SYSTEM
Parameter Recommended Tolerance
Mass spectrometer
Data system
Scan range
Scan time
Resolution
Ion source temperature
Electron energy
Trap current
Multiplier voltage
Preamplifier sensitivity
Finnigan 4023
Incos 2400
95-550
1 sec
Unit
280°C
70 eV
0.2 mA
-1,600 V
10"6 A/V
Other3
Other
Other
Otherb
Optimum performance
200°-300°C
Optimum performance
Optimum performance
Optimum performance
Set for desired
working range
a Substitutions permitted if performance criteria are met.
b Greater than five data points over a GC peak is a minimum.
c Filaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no separator is used.
C-22
-------
TABLE 8. OPERATING PARAMETERS FOR MAGNETIC SECTOR MASS SPECTROMETER SYSTEM
Parameter
Mass spectrometer
Data system
Scan range
Scan mode
Cycle time
Resolution
Ion source temperature
Electron energy
Emission current
Filament current
Multiplier
Recommended
Finnigan MAT 311A
Incos 2400
98-550
Exponential
1.2 sec
1,000
280°C
70 eV
1-2 mA
Optimum
-1,600 V
Tolerance
Other3
Other
Other
Other
Otherb
> 500
250-300°
70 eV
Optimum
Optimum
Optimum
a Substitutions permitted if performance criteria are met.
b Greater than five data pointy over a GC peak is a minimum.
c Filaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no separator is used.
C-23
-------
7.4 The PCB response factors (RF ) must be determined using Equation
7-1 for the analyte homologs?
A x M.
RF = v£ ^ Eq. 7-1
p A. x M ^
r is p
where RF = response factor of a given PCB isomer
A = area of the characteristic ion for the PCB congener
" peak
M = mass of PCB congener injected (nanograms)
A. = area of the characteristic ion for the internal
standard peak
M. = mass of internal standard injected (nanograms)
X. S
If specific congeners are known to be present and if standards
are available, selected RF values may be employed. For general
samples, solutions CSxxx and SSxxx or a mixture (Tables 3 and 4)
may be used as the response factor solution. The PCB-surrogate
pairs to be used in the RF calculation are listed in Table 9.
Generally, only the primary ions of both the analyte and surrogate
are used to determine the RF values. If alternate ions are to be
used in the quantitation, the RF must be determined using that
characteristic ion.
The RF value must be determined in a manner to assure ±20% accu-
racy and precision. For instruments with good day-to-day preci-
sion, a running mean (RF) based on seven values determined once
each day may be appropriate. Other options include, but are not
limited to, triplicate determinations of a single concentration
spaced throughout a day or determination of the RF at three dif-
ferent levels to establish a working curve.
If replicate RF values differ by greater than ±10% RSD, the system
performance should be monitored closely. If the RSD is greater
than ±20%, the data set must be considered invalid and the RF re-
determined before further analyses are done.
7.5 If the GC/EIMS system has not been demonstrated to yield a linear
response or if the analyte concentrations are more than one order
of magnitude different from those in the RF solution, a calibra-
tion curve must be prepared. If the analyte and RF solution con-
centrations differ by more than one order of magnitude, a calibra-
tion curve should be prepared. A calibration curve should be
established with triplicate determinations at three or more con-
centrations bracketing the analyte levels.
C-24
-------
TABLE 9. PAIRINGS OF ANALYTE, CALIBRATION. AND SURROGATE COMPOUNDS
Analyte
Congener
no . Compound
Calibration standard
Congener
no.
1 2-Ci2H9Cl 1
2,3 3- and 4-C12H9Cl 3
4-15 C12HaCl2 7
o
10
i c. on f* u r*~\
ID™ Jy LI 2X170-1.3
40-81 C12H6C14
82-127 C12H5C15
128-169 C12H4C16
•f T/\ "f f\ O f* TT f* 1
i/U-iyo Li2n3Lil7
monc r> u PI
-/Ub Li2n2Llg
206—208 Ci2HCl9
209 Ci2Clio
30
50
97
143
183
202
207
209
2
4
2
2
2
2
2
?
?,
7
C
,4
,4,
,2'
,2'
,2'
,2'
,2'
?'
12C
6
,4
,3
,3
,3
,3
3
ll
Compound
,6
',4,
,4,5
',4,
3'
0
5
,6'
4' ,5', 6
5, 5', 6, 6'
4, 4', 5, 6, 6'
Surrogate
Congener
no.
211
211
211
212
212
212
212
213
213
213
214
Compound
J3C6-4
13Cg-4
13C12-3,3'
13Ci2-3,3'
13Ci2-3,3'
1 3p _ o o t
13C12-2,2'
13C12-2,2'
13C12-2,2'
13Ci2Cll0
,4,4'
,4,4'
,4,4'
,4,4'
,3, 3', 5, 5'
,3, 3', 5, 5'
,3,3', 5, 5'
,6,6'
,6,6'
,6,6'
a Ballschmiter numbering system, see Table 1.
-------
7.6 The relative retention time (RRT) windows for the 10 horaologs and
surrogates must be determined. If all congeners are not available,
a mixture of available congeners or an Aroclor mixture (e.g.,
1016/1254/1260) may be used to estimate the windows. The windows
must be set wider than observed if all isomers are not determined.
Typical RRT windows for one column are listed in Table 10. The
windows may differ substantially if other GC parameters are used.
8.0 Sample Collection, Handling, and Preservation
The sampling shall be conducted by competent personnel experienced with
this test procedure and cognizant of the constraints of the anaytical
techniques for PCBs, particularly contamination problems.
8.1 Stack sampling1
8.1.1 Pretest preparation - All train components shall be main-
tained and calibrated according to the procedure de-
scribed in APTD-0581,3 unless otherwise specified herein.
This should be done in the laboratory prior to sampling.
8.1.1.1 Cleaning glassware - All glass parts of the
train upstream of and including the adsorbent
tube and impingers, should be cleaned as de-
scribed in Section 3.1.1. Special care should
be devoted to the removal of residual silicone
grease sealants, on ground glass connections of
used glassware. These grease residues should
be removed by soaking several hours in a chromic
acid cleaning solution prior to routine cleaning
as described above.
8.1.1.2 Solid adsorbent tube - 7.5 g of Florisil acti-
vated within the last 30 days and still warm
from storage in a 110°C oven, is weighed into
the adsorbent tube (prerinsed with hexane) with
a glass wool plug in the downstream end. A
second glass wool plug is placed in the tube to
hold the sorbent in the tube. Both ends of the
tube are capped with ground glass caps. These
caps should not be removed until the tube is
fitted to the train immediately prior to sampling.
8.1.2 Preliminary determinations - The sampling site and the
minimum number of sampling points are selected according
to Method I2 or as specified by the Agency. The stack
pressure, temperature, and the range of velocity heads
are determined using Method 22 and moisture content using
Approximation Method 42 or its alternatives for the pur-
pose of making isokinetic sampling rate calculations.
Estimates may be used. However, final results must be
based on actual measurements made during the test.
C-26
-------
TABLE 10. RELATIVE RETENTION TIME (RRT) RANGES OF PCB HOMOLOGS
VERSUS d6-3.3'.4,4'-TETRACHLOROBIPHENYL
No. of
PCB isomers
homolog measured
Monochloro
Dichloro
Trichloro
Tetrachloro
Pentachloro
Hexachloro
Heptachloro
Octachloro
Nonachloro
Decachloro
3
10
9
16
12
13
4
6
3
1
Observed range
of RRTsa
0.40-0.50
0.52-0.69
0.62-0.79
0.72-1.01
0.82-1.08
0.93-1.20
1.09-1.30
1.19-1.36
1.31-1.42
1.44-1.45
Congener
no.
1
3
7
30
50
97
143
183
202
207
209
Observed
RRTa
0.43
0.50
0.58
0.65
0.75
0.98
1.05
1.15
1.19
1.33
1.44
Projected
range of
RRTs
0.35-0.55
0.35-0.80
0.35-0.10
0.55-1.05
0.80-1.10
0.90-1.25
1.05-1.35
1.10-1.50
1.25-1.50
1.35-1.50
a The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were
measured versus 3,3',4,4'-tetrachlorobiphenyl-de (internal standard) using
a 15-m J&W DB-5 fused silica column with a temperature program of 110°C
for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an on-
column injector. A Finnigan 4023 Incos quadrupole mass spectrometer oper-
ating with a scan range of 95-550 daltons was used to detect each PCB
congener.
b The projected relative retention windows account for overlap of eluting
homologs and take into consideration differences in operating systems and
lack of all possible 209 PCB congeners.
C-27
-------
The molecular weight of the stack gases is determined
using Method 3.2
A nozzle size is selected based on the maximum velocity
head so that isokinetic sampling can be maintained at a
rate less than 0.75 cfm. It is not necessary to change
the nozzle size in order to maintain isokinetic sampling
rates. During the run, the nozzle size must not be
changed.
A suitable probe length is selected such that all traverse
points can be sampled. Sampling from opposite sides for
large stacks may be considered to reduce the length of
probes.
A sampling time is selected appropriate for total method
sensitivity and the PCB concentration anticipated. Sam-
pling times should generally fall within a range of 2 to
4 hr.
A buzzer-timer should be incorporated in the control box
(see Figure 1) to alarm the operator to move the probe to
the next sampling point.
8.1.3 Preparation of collection train - During preparation and
assembly of the sampling train, all train openings must
be covered until just prior to assembly or until sampling
is about to begin. Immediately prior to assembly, all
parts of the train upstream of the adsorbent tube are
rinsed with hexane. The probe is marked with heat resis-
tant tape or by some other method at points indicating
the proper distance into the stack or duct for each sam-
pling point.
200 ml of water is placed in each of the first two impin-
gers, and the third impinger left empty. CAUTION: Sealant
greases must not be used in assembling the train. If the
preliminary moisture determination shows that the stack
gases are saturated or supersaturated, one or two addi-
tional empty impingers should be added to the train be-
tween the third impinger and the Florisil tube. See
Section 5.1.5. Approximately 200 to 300 g or more, if
necessary, of silica gel is placed in the last impinger.
Each impinger (stem included) is weighed and the weights
recorded to the nearest 0.1 g on the impingers and on
the data sheet.
Unless otherwise specified by the Agency, a temperature
probe is attached to the metal sheath of the sampling
probe so that the sensor is at least 2.5 cm behind the
nozzle and pitot tube and does not touch any metal.
C-28
-------
The train is assembled as shown in Figure 1. Through all
parts of this method use of sealant greases such as stop-
cock grease to seal ground glass joints must be avoided.
Crushed ice is placed around the impingers.
8.1.4 Leak check procedure - After the sampling train has been
assembled, the probe heating system(s) is turned on and
set (if applicable) to reach a temperature sufficient to
avoid condensation in the probe. Time is allowed for the
temperature to stabilize. The train is leak checked at
the sampling site by plugging the nozzle and pulling a
380 mm Hg (15 in. Hg) vacuum. A leakage rate in excess
of 4% of the average sampling rate or 0.0057 m3/min
(0.02 cfm) whichever is less, is unacceptable.
The following leak check instruction for the sampling
train described in APTD-05813 may be helpful. The pump
is started with bypass valve fully open and coarse adjust
valve completely closed. The coarse adjust valve is
partially opened and the bypass valve slowly closed until
380 mm Hg (15 in. Hg) vacuum is reached. The direction
of bypass valve must not be reversed. This will cause
water to back up into the probe. If 380 mm Hg (15.in. Hg)
is exceeded, either the leak check is conducted at this
higher vacuum or the leak check is ended as described
below and start over.
When the leak check is completed, the plug is first slowly
removed from the inlet to the probe and the vacuum pump
is immediately turned off. This prevents the water in
the impingers from being forced backward into the probe.
Leak checks, shall be conducted as described above prior
to each test run and at the completion of each test run.
If leaks are found to be in excess of the acceptable rate,
the test will be considered invalid. To reduce lost time
due to leakage occurrences, it is recommended that leak
checks be conducted between port changes.
8.1.5 Train operation - During the sampling run, an isokinetic
sampling rate within 10%, or as specified by the Agency,
of true isokinetic shall be maintained. During the run,
the nozzle or any other part of the train in front of
and including the Florisil tube must not be changed.
For each run, the data required on the data sheets must
be recorded. An example is shown in Figure 4. The dry
gas meter readings are recorded at the beginning and end
of each sampling time increment, when changes in flow
rates are made, and when sampling is halted. Other data
point readings are taken at least once at each sample
point during each time increment and whenever significant
C-29
-------
FIELD DATA
PLANT.
DATE._
SAMPLING LOCATION .
SAMPLE TYPE
RUN NUMBER.
OPERATOR
PROBE LENGTH AND TYPE.
NOZZLE I.D..,..
ASSUMED MOISTURE. °,
AMBIENT TEMPERATURE
BAROMETRIC PRESSURE .
STATIC PRESSURE. (Ps)_
FILTER NUMBER (s)
SAMPLE BOX NUMBER
METER BOX NUMBER
METER AHf,
C FACTOR
PROBE HEATER SETTING
HEATER BOX SETIING_
REFERENCE Ap
SCHEMATIC OF TRAVERSE POINT LAYOUT
READ AND RECORD ALL DATA EVERY.
MINUTES
TRAVERSE
POINT
NUMBER
\. CLOCK TIME
™fLING XnS™
TIMt.mm N^
__
GAS METER READING
(Vm). It3
VELOCITY
HEAD
(&PS). in. H?0
ORIFICE PRESSURE
DIFFERENTIAL
(AH), in. HjOl
DESIRED
ACTUAL
STACK
TEMPERATURE
."F
DRY GAS METER
TEMPERATURE
INLET
-------
changes (20% variation in velocity head readings) neces-
sitate additional adjustments in flow rate.
The portholes are cleaned prior to the test run to mini-
mize change of sampling deposited material. To begin
sampling, the nozzle cap is removed, the probe heater
operational and temperature up, and the pitot tube and
probe positions are verified (if applicable). The nozzle
is positioned at the first traverse point with the tip
pointing directly into the gas stream. The pump is
started and the flow adjusted to isokinetic conditions.
Nomographs are available for sampling trains using type
S pitot tubes with 0.85 ± 0.02 coefficients (C ), and
when sampling in air or a stack gas with equivalent
density (molecular weight, M,, equal to 29 ± 4), which
aid in the rapid adjustment of the isokinetic sampling
rate without excessive computations. If C and M, are
outside the above stated ranges, the nomograph cannot be
used unless appropriate steps are taken to compensate for
the deviations.
When the stack is under significant negative pressure
(height of impinger stem), the coarse adjust valve must
be closed before inserting the probe into the stack to
avoid water backing into the probe. If necessary, the
pump may be turned on with the coarse valve closed.
When the probe is in position, the openings around the
probe and porthole must be blocked off to prevent un-
representative dilution of the gas stream.
The stack cross section is traversed, as required by
Method I2 or as specified by the Agency. To minimize
chance of extracting deposited material, the probe nozzle
should not bump into the stack walls when sampling near
the walls or when removing or inserting the probe through
the portholes.
During the test run, periodic adjustments are made to
keep the probe temperature at the proper value. More
ice and, if necessary, salt is added to the ice bath to
maintain a temperature of less than 20°C (68°F) at the
impinger/silica gel outlet, to avoid excessive moisture
losses. Also, the level and zero of the manometer should
be periodically checked.
If the pressure drop across the train becomes high enough
to make isokinetic sampling difficult to maintain, the
test run should be terminated. Under no circumstances
should the train be disassembled during the test run to
determine and correct causes of excessive pressure drops.
C-31
-------
At the end of the sample run, the pump is turned off, the
probe and nozzle removed from the stack, and the final
dry gas meter reading recorded. A leak check is performed,
with acceptability of the test run based on the same cri-
teria as in Section 8.1.4. The percent isokinetic is
calculated (see calculation section) to determine whether
another test run should be made. If there is difficulty
in maintaining isokinetic rates due to source conditions,
the Agency should be consulted for possible variance on
the isokinetic rates.
8.1.6 Blank train - For each series of test runs, a blank train
is set up in a manner identical to that described above,
but with the nozzle capped with aluminum foil and the
exit end of the last impinger capped with a ground glass
cap. The train is allowed to remain assembled for a
period equivalent to one test run. The blank sample is
recovered as described in Section 8.3.
8.2 Static air sampling3 - The sampling procedure for static air is
identical to that described in Section 8.1 with the following ex-
ceptions: (a) impingers and a beatable probe are not required
prior to the adsorbent tube; and (b) the PCB concentrations may
dictate a longer or shorter sampling time.
The selection of sampling time and rate should be based on the
approximate levels of PCB residues expected in the sample. The
sampling rate should not exceed 14 liter/min and may typically
fall in the range of 5 to 10 liter/min. Sampling times should be
more than 20 min but should not exceed 4 hr.
8.3 Sample recovery - Proper cleanup procedure begins as soon as the
probe is removed from the stack at the end of the sampling period.
When the probe can be safely handled, all external particulate
matter near the tip of the probe nozzle is wiped off. The probe
is removed from the train and both ends closed off with aluminum
foil. The inlet to the train is capped off with a ground glass
cap.
The probe and impinger assembly are transfered to the cleanup area.
This* area should be clean and protected from the wind so that the
chances of contaminating or losing the sample will be minimized.
The train is inspected prior to and during disassembly and any
abnormal conditions noted. The samples are treated as follows:
8.3.1 Adsorbent tube - The Florisil tube is removed from the
train and capped with ground glass caps.
8.3.2 Sample Container No. 1 - The first three impingers are
removed. The outside of each impinger is wiped off to
remove excessive water and other debris. The impingers
C-32
-------
are weighed (stem included), and the weight recorded on
a data sheet. The contents are poured directly into
Container No. 1.
8.3.3 Sample Container No. 2 - Each of the first three impingers
are rinsed sequentially with 30-ml acetone and then with
30-ml hexane, and the rinses put into Container No. 2.
Material deposited in the probe is quantitatively recov-
ered using 100-ml acetone and then 100-ml hexane and
these rinses added to Container No. 2.
8.3.4 Silica gel container - The last impinger is removed, and
the outside wiped to remove excessive water and other
debris. It is weighed (stem included), and the weight
recorded on the data sheet. The contents are transferred
to the used silica gel can.
8.4 Sample preservation - Samples should be stored in the dark at 4°C.
Storage times in excess of 4 weeks are not recommended.
9.0 Sample Preparation1
9.1 Extraction
9.1.1 Adsorbent tube - The entire contents of the adsorbent
tube are expelled directly onto a glass wool plug in the
sample holder of a Soxhlet extractor. Although no extrac-
tion thimble is required, a glass thimble with a coarse-
fritted bottom may be used.
The tube is rinsed with 5-ml acetone and then with 15-ml
hexane and these rinses put into the extractor. The ex-
traction apparatus is assembled and the adsorbent ex-
tracted with 170-ml hexane for at least 4 hr. The ex-
tractor should cycle 10 to 14 times per hour. After
allowing the extraction apparatus to cool to ambient
temperature, the extract is transferred into a Kuderna-
Danish evaporator.
The extract is evaporated to about 5 ml on a steam bath
and the evaporator allowed to cool to ambient temperature
before disassembly. The extract is transferred to a 50-ml
separatory funnel and the funnel set aside.
9.1.2 Sample Container No. 1 - The aqueous sample is transferred
to a 1,000-ml separatory funnel. The container is rinsed
with 20-ml acetone and then with two 20-ml portions of
hexane, adding the rinses to the separatory funnel.
The sample is extracted with three 100 ml portions of
hexane and the sequential extracts transferred to a
Kuderna-Danish evaporator.
C-33
-------
The extract is concentrated to about 5 ml and allowed to
cool to ambient temperature before disassembly. The ex-
tract is filtered through a micro column of anhydrous
sodium sulfate into a 50-ml separatory funnel containing
the corresponding Florisil extract from Section 9.1.1.
The micro column is prepared by placing a small plug of
glass wool in the bottom of the large portion of a dis-
posable pipette and then adding anhydrous sodium sulfate
until the tube is about half full.
9.1.3 Sample Container No. 2 - The organic solution is trans-
ferred into a 1,000-ml separatory funnel. The container
is rinsed with two '20 ml portions of hexane and the rinses
added to the separatory funnel. The sample is washed with
three 100 ml portions of water. The aqueous layer is
discarded and the organic layer transferred to a Kuderna-
Danish evaporator.
The extract is concentrated to about 5 ml and allowed to
cool to ambient temperature before disassembly. The ex-
tract is filtered through a micro column of anhydrous
sodium sulfate into the 50-ml separatory funnel contain-
ing the corresponding Florisil and impinger extracts
(Section 9.1.2).
9.2 Cleanup - Two tested cleanup techniques are described below.4 De-
pending upon the complexity of the sample, one or both of the tech-
niques may be required to fractionate the PCBs from interferences.
If the sample extract is colored, the Florisil column cleanup may
be indicated.
9.2.1 Acid cleanup
9.2.1.1 Add 5 ml of concentrated sulfuric acid to the
separatory funnel containing the sample extract
and shake for 1 min.
9.2.1.2 Allow the phases to separate, transfer the
sample (upper phase) with three 1 to 2 ml
solvent rinses to Kuderna-Danish evaporator
and concentrate to an appropriate volume.
9.2.1.3 Analyze as described in Section 10.0.
9.2.1.4 If the sample is highly contaminated, a second
or third acid cleanup may be employed.
9.2.2 Florisil column cleanup
9.2.2.1 Variations among batches of Florisil may affect
the elution volume of the various PCBs. For
this reason, the volume of solvent required to
C-34
-------
completely elute all of the PCBs must be veri-
fied by the analyst. The weight of Florisil
can then be adjusted accordingly.
9.2.2.2 Place a 20-g charge of Florisil, activated over-
night at 130°C, into a Chromaflex column. Settle
the Florisil by tapping the column. Add about
1 cm of anhydrous sodium sulfate to the top of
the Florisil. Pre-elute the column with 70-80
ml of hexane. Just before the exposure of the
sodium sulfate layer to air, stop the flow.
Discard the eluate.
9.2.2.3 Add the sample extract to the column. Add 225
ml of hexane to the column. Carefully wash
down the inner wall of the column with a small
amount of the hexane prior to adding the total
volume. Discard the first 25 ml.
9.2.2.A Collect 200 ml of hexane eluate in a Kuderna-
Danish flask. All of the PCBs should be in
this fraction. Concentrate to an appropriate
volume.
9.2.2.5 Analyze the sample as described in Section 10.0.
10.0 Gas Chromatographic/Electron Impact Mass Spectrometric Determination
10.1 Internal standard addition - Pipet an appropriate volume of inter-
nal standard solution SSxxx into the sample. The final concentra-
tion of the internal standards must be in the working range of the
calibration and well above the matrix background. The internal
standards are thoroughly mixed by mechanical agitation.
Note: The volume measurement of the spiking solution is critical
to the overall method precision. The analyst must exercise cau-
tion that the volume is known ±1% or better. Where necessary,
calibration of the pipet is recommended.
Note: This same solution is used as a surrogate standard solution
in the protocols for products/product waste and for water. In
this protocol, the 13C-labeled PCBs are spiked after extraction,
so are used as internal standards.
Alternately, another internal standard solution such as the dg-
3,3',4,4'-tetrachlorobiphenyl used in the product/product waste
and water protocols may be used, if acceptable RF precision and
accuracy are shown across the homolog range.
i
10.2 Tables 2, and 5 through 8 summarize the recommended operating con-
ditions for analysis. Figure 5 presents an example of a chromato-
gram.
C-35
-------
iet).o-i
me
o
i
LO
en cu
4-J
,H CO
cfl 00
C 0
M r-l
CU r-l
4J 3
C "I
i- ^
vO vO CO
O 13 rH
rH rH rH
1 X X
rH rH . C C
XX CU CU
c c x; x;
(U CU P. p.
Xi XI "r-t "H
P. P. Xi X)
•r( -H O O
Xl XI ^~^ r-llfi.V'jH
O O CU o
M VJ J-l rH
o o CD x;
rH rH 00 O
Xi Xi O cfl
O O r-l r-*
O O M rH 4-1
C C 3 X cu
O O W C H
a s ^ cu i
II rC
rH H Xi -
rH
£>-,
C
CU
x:
P.
•H
Xi
o
^i
o
rH
Xi
o
o
C
o
^
1
CN
018
|
X O P. CO
C rH -H
rH cu x; - xi co
x xi o o
C P, Cfl rJ
CU -H r-l O
Xi XI U rH
P. O (U X!
•H M H a
XI O I cfl
O rH VO -U
MX:- c
o a
ft ^ t
T T
L L I « II . ». - 11.
rH
x;
o
CO rH
V-i X
•u C
a» cu
H x:
i ex
•H
^J" XI
« o
-^J- ^
- o
- rH
co x;
•- o
PO tfl
rN
CU
pe
W
vD
«
m
r.
^-
A
^J-
*
CO
tl
rH
rH C
X cu
C JC
CU p.
Xi -H rH
P. X) X
•H 0 C
XI MO)
o ox;
r-l rH P.
o x: -H
rH O XI
X! CO 0
O C r-l
CO 0 O
4-1 K rH
a i x!
O 0
1 ^O Cfl
o
vO v£> (1)
Q
vO iO
rv r.
iO <3"
*. *%
•O ••^"
f. ft
+. —
ro ro
CO CO
ft ft
*. —
CN CN
r. ft
CN OJ
^N .200 K °
•1
CN
II7U
12."
20:
,»„ _ ^Jl
"
' 112592.
CN
1 — 1
U
CO
rH
1
rH
C
CU
x;
P, ,—>
•H CU
f"1 4J
O cfl
M Ml
0 0
i — 1 M
Xi r-l
O 3
cfl o)
o ^
CU
Q
_H
3 I'.'f'j SCA:I
C!) 23:20 1 1 III i
Figure 5. Capillary gas chromatography/electron impact ionization mass spectrometry (CGC/EIMS)
chromatogram or the calibration standard solution required for quantitatlon of PCBs by homolog.
This chromatogram includes PCBs representative of each homolog, three carbon-13 labeled surrogates,
and the deuterated internal standard. The concentration of all components and the CGC/EIMS
parameters are presented in 'Cables 3, 4, 5, and 7.
-------
10.3 While the highest available chromatographic resolution is not a
necessary objective of this protocol, good chromatographic per-
formance is recommended. With the high resolution of CGC, the
probability that the chromatographic peaks consist of single
compounds is higher than with PGC. Thus, qualitative and quanti-
tative data reduction should be more reliable.
10.4 After performance of the system has been certified for the day
and all instrument conditions set according to Tables 2, and 5
through 8, inject an aliquot of the sample onto the GC column.
If the response for any ion, including surrogates and internal
standard, exceeds the working range of the system, dilute the
sample and reanalyze. If the responses of surrogates, internal
standard, or analytes are below the working range, recheck the
system performance. If necessary, concentrate the sample and
reanalyze.
10.5 Record all data on a digital storage device (magnetic disk, tape,
etc.) for qualitative and quantitative data reduction as discussed
below.
11.0 Qualitative Identification
11.1 Selected ion monitoring (SIM) or limited mass scan (IMS) data -
The identification of a compound as a given PCB homolog requires
that two criteria be met:
11.1.1 (1) The peak must elute within the retention time window
set for that homolog (Section 7.6); and (2) the ratio of
two ions obtained by SIM (Table 11) or by IMS (Table 12)
must match the natural ratio within ±20%. The analyst
must search the higher mass windows, in particular M+70,
to prevent misidentification of a PCB fragment ion clus-
ter as the parent.
11.1.2 If one or the other of these criteria is not met, inter-
ferences may have affected the results and a reanalysis
using full scan EIMS conditions is recommended.
11.2 Full scan data
11.2.1 The peak must elute within the retention time windows
set for that homolog (as described in Section 7.6).
11.2.2 The unknown spectrum must match that of an authentic PCB.
The intensity of the three largest ions in the molecular
cluster (two largest for monochlorobiphenyls) must match
the natural ratio within ±20%. Frequent clusters with
proper intensity ratios must also be present.
11.2.3 Alternatively, a spectral search may be used to auto-
matically reduce the data. The criteria for acceptable
C-37
-------
TABLE 11. CHARACTERISTIC SIM IONS FOR PCBs
Homolog
Ci2H9Cl
Ci2H8Cl2
C12H7C13
C12H6C14
Ci2H5Cl5
C12H4C16
C^HgCly
C12H2C18
Ci2HClg
Ci2Clio
Primary
188 (100)
222 (100)
256 (100)
292 (100)
326 (100)
360 (100)
394 (100)
430 (100)
464 (100)
498 (100)
Ion (relative intensity)
Secondary
190 (33)
224 (66)
258 (99)
290 (76)
328 (66)
362 (82)
396 (98)
432 (66)
466 (76)
500 (87)
Tertiary
-
226 (11)
260 (33)
294 (49)
324 (61)
364 (36)
398 (54)
428 (87)
462 (76)
496 (68)
Source: Rote, J. W., and W. J. Morris, "Use of Isotopic Abundance Ratios in
Identification of Polychlorinated Biphenyls by Mass Spectrometry,"
J. Assoc. Offic. Anal. Chem., 56(1), 188-199 (1973).
C-38
-------
TABLE 12. LIMITED MASS SCANNING (IMS) RANGES FOR PCBs
Compound
C^Cl,
C12H8C12
C12H7C13
C12H6C13
C^H5C15
Ci2H4Cl6
C12H3C17
C12H2Clg
Ci2HCl9
C12Cl!0
Ci2D6Cl4
13C612C6H9C1
13C12H6C14
13C12H2C18
13C12C116
Mass range (m/z)
186-190
220-226
254-260
288-294
322-328
356-364
386-400
426-434
460-468
494-504
294-300
192-196
300-306
438-446
506-516
a Adapted from Tindall, G. W., and P. E. Wininger, "Gas Chromatography-Mass
Spectrometry Method for Identifying and Determining Polychlorinated Bi-
phenyls," J. Chromatogr., 196, 109-119 (1980).
C-39
-------
identification include a high index of similarity. For
the Incos 2300, a fit of 750 or greater must be obtained.
11.3 Disputes in interpretation - Where there is reasonable doubt as
to the identity of a peak as a PCB, the analyst must either iden-
tify the peak as a PCB or proceed to a confirmational analysis
(see Section 13.0).
12.0 Quantitative Data Reduction
12.1 Once a chromatographic peak has been identified as a PCB, the com-
pound is quantitated based either on the integrated abundance of
the SIM data or EICP for the primary characteristic ion in Tables
11 and 12. If interferences are observed for the primary ion, use
the secondary and then tertiary ion for quantitation. If inter-
ferences in the parent cluster prevent quantitation, an ion from a
fragment cluster (e.g., M-70) may be used. Whichever ion is used,
the RF must be determined using that ion. The same criteria
should be applied to the internal standard compounds (Table 13).
12.2 Using the appropriate response factor (RF ) as determined in Sec-
tion 7.3, calculate the mass of each PCB peak (M ) using Equation
12-1. p
A -
M = _E . i . M
p A. RF is Eq. 12-1
* is p ^
where A = area of the characteristic ion for the analyte PCB
^ peak
A. = area of the characteristic ion for the internal
standard peak
RF = response factor of a given PCB congener
M. = mass of internal standard injected (micrograms)
1S
12.3 If a peak appears to contain non-PCB interferences which cannot
be circumvented by a secondary or tertiary ion, either:
12.3.1 Reanalyze the sample on a different column which sepa-
rates the PCB and interferents;
12.3.2 Perform additional chemical cleanup (Section 9) and then
reanalyze the sample; or
12.3.3 Quantitate the entire peak as PCB.
12.4 Sum all of the peaks for each homolog and then sum those to yield
the total PCB mass, M™, in the sample. If a concentration-per-
peak or concentration-per-homolog reporting format is desired,
carry each value through the calculations in an appropriate manner.
C-40
-------
TABLE 13. CHARACTERISTIC IONS FOR 13C-LABELED PCB SURROGATES
Ion (relative intensity)
Specific compound Primary Secondary Tertiary
13C612C6H9C1 194 (100) 196 (33)
13C12H6C14 304 (100) 306 (49) 302 (78)
13C12H2C18 442 (100) 444 (65) 440 (89)
13C12C110 510 (100) 512 (87) 514 (50)
C-41
-------
12.5 Calculation of air sample volume1
12.5.1 Nomenclature
M = Mass of PCB represented by a chromatographic peak
" micrograms
= Total mass of PCBs in sample, micrograms
C = Concentration of PCBs in air, micrograms per cubic
meter, corrected to standard conditions of 20°C,
760 mm Hg (68°F, 29.92 in. Hg) on dry basis
A = Cross-sectional area of nozzle, square meter (square
n feet)
B = Water vapor in the gas stream, proportion by volume
I = Percent of isokinetic sampling
MW = Molecular weight of water, 18 g/g-mole (18 lb/
Ib-mole)
P, = Barometric pressure at the sampling site, mm Hg
oar / . YT \
(in. Hg)
P = Absolute stack gas pressure, mm Hg (in. Hg)
s
P , = Standard absolute pressure, 760 mm Hg (29.92 in
SL.U TT \
Hg)
R = Ideal gas constant, 0.06236 mm Hg-m3/K-g-mole (21.83 in.
Hg-ft3/°R-lb-mole)
T = Absolute average dry gas meter temperature °K (°R)
T = Absolute average stack gas temperature °K (°R)
s
= Standard absolute temperature, 293°K (528°R)
V, = Total volume of liquid collected in impingers and
silica gel, milliliters. Volume of water col-
lected equals the weight increase in grams times
1 ml/g
V = Volume of gas sample as measured by dry gas meter,
ra dcm (dcf)
V , ,.. = Volume of gas sample measured by the dry gas
meter corrected to standard conditions,
dscm (dscf)
C-42
-------
V , ,x = Volume of water vapor in the gas sample cor-
^ rected to standard conditions, scm (scf)
V = Total volume of sample, railliliter
V = Stack gas velocity, calculated by EPA Method 2,
m/sec (ft/sec)
AH = Average pressure differential across the orifice
meter, mm H20 (in.
p = Density of water, 1 g/ml (0.00220 Ib/ml)
W
9 = Total sampling time, minutes
13.6 = Specific gravity of mercury
60 = Seconds per minute
100 = Conversion to percent
12.5.2 Average dry gas meter temperature and average orifice
pressure drop - See data sheet (Figure 4).
12.5.3 Dry gas volume - Correct the sample volume measured by
the dry gas meter to standard conditions [20°C, 760 mm Hg
(68°F, 29.92 in. Hg)} by using Equation 12-2.
P, +™ P,
w v bar 13.6 „ 17 *bar 13.6 Eq. 12-2
m(std) " m ~T~ P~T" K ra T
^ J m std m
where K = 0.3855°K/mm Hg for metric units
= 17.65 °R/in. Hg for English units
12.5.4 Volume of water vapor
= vic ar ?rf = KVic E*- 12'3
w std
where K = 0.00134 m3/ml for metric units
= 0.0472 ft3/ml for English units
12.5.5 Moisture content
R - w(std) ,
Bws - V , _, + V f ¥.. Eq" 12"4
m(std) w(std)
If the liquid droplets are present in the gas stream, as-
sume the stream to be saturated and use a psychrometric
chart to obtain an approximation of the moisture per-
centage.
C-43
-------
12.6 Concentration of PCBs in stack gas - Determine the concentration
of PCBs in the air according to Equation 12-5 and report in micro-
grams per cubic meter using Table 14. If an alternate reporting
format (e.g., concentration per peak) is desired, a different
report form may be used.
M
C = K ^ — - - Eq. 12-5
a m(std)
where K = 35.31 ft3/m3
12.7 Isokinetic variation
12.7.1 Calculations from raw data.
T 100 Ts [K Vlc + (VTm) (Pbar) + M/13'6^ _ „,
i = - 6o e v p A - — Eq- 12'6
s s n
where K = 0.00346 mm Hg-m3/ml-°K for metric units
= 0.00267 in. Hg-ft3/ml-°R for English units
12.7.2 Calculations from intermediate values
T V , „ ,, P . , 100
- s m(std) std _ _ _
~
T ^ . V 0 A P 60 (1-B )
std s n s ws
T V
v s m(std)
— K. ---- ' -------
P V A 0 (1-B )
s s n ws
where K = 4.323 for metric units
= 0.0944 for English units
12.7.3 Acceptable results - The following range sets the limit
on acceptable isokinetic sampling results:
If 90% < I < 110%, the results are acceptable. If the
results are low in comparison to the standards and I is
beyond the acceptable range, the Agency may opt to ac-
cept the results.
12.8 Round off all numbers reported to two significant figures.
13.0 Confirmation
If there is reason to question the qualitative identification (Section
11.0), the analyst may choose to confirm that a peak is not a PCB. Any
technique may be chosen provided that it is validated as having equiva-
lent or superior selectivity and sensitivity to GC/EIMS. Some candidate
techniques include alternate GC columns (with EIMS detection), GC/CIMS,
GC/NCIMS, high resolution EIMS, and MS/MS techniques. Each laboratory
C-44
-------
TABLE 14. ANALYSIS REPORT
Sample No.
Sample Matrix
Sample Source
INCIDENTAL PCBs IN AIR
Notebook No. or File Location
Volume Collected [V , ,-J
Mass of Internal Standard
Analyte 1° 2° Il°
IS 298 246
1-C1 188 190
2-C1 222 224
3-C1 256 258
4-C1 292 290
5-C1 326 328
6-C1 360 362
7-C1 394 396
8-C1 430 432
9-C1 464 466
10-C1 498 500
Total (M^
Concentration (C.)
A
Reported by:
Name
Signature/Date
Organization
m3
Injected, M._
Qualitative
I2° Ratio Theoretical
100/76
100/33
100/66
100/99
100/76
100/66
100/82
100/98
100/66
100/76
100/87
Internal Audit:
Name
Signature/Date
Organization
M8
Quantitative
Ion Mass
OK? Used RF M (pg)
P
1.000
M8
jjg/m3
EPA Audit:
Name
Signature/Date
Organization
C-45
-------
must validate confirmation techniques to show equivalent or superior
selectivity between PCBs and interferences and sensitivity (limit of
quantitation, LOQ).
If a peak is confirmed as being a non-PCB, it may be deleted from the
calculation (Section 12). If a peak is confirmed as containing both
PCB and non-PCB components, it must be quantitated according to Section
12.3.
14.0 Quality Control
14.1 Each laboratory that uses this method must operate a formal qual-
ity control (QC) program. The minimum requirements of this pro-
gram consist of an initial demonstration of laboratory capability
and the analysis of spiked samples as a continuing check on per-
formance. The laboratory must maintain performance records to
define the quality of data that are generated. After a date spe-
cified by the Agency, ongoing performance checks should be com-
pared with established performance criteria to determine if the
results of analyses are within accuracy and precision limits ex-
pected of the method.
14.2 The analysts must certify that the precision and accuracy of the
analytical results are acceptable by:
14.2.1 The absolute precision of surrogate recovery, measured
as the RSD of the integrated EIMS area (A ) for a set
of samples, must be ±10%.
14.2.2 The mean recovery (R ) of at least four replicates of a
QC check sample to be supplied by the Agency must meet
Agency-specified accuracy and precision criteria. This
forms the initial data base for establishing control
limits (see Section 14.3 below).
14.3 Control limits - The laboratory must establish control limits using
the following equations:
Upper control limit (UCL) = R + 3 RSD
Upper warning limit (UWL) = R + 2 RSD
Lower warning limit (LWL) = R - 2 RSD
Lower control limit (LCL) = R - 3 RSD
c c
These may be plotted on control charts. If an analysis of a check
sample falls outside the warning limits, the analyst should be
alerted that potential problems may need correction. If the re-
sults for a check sample fall outside the control limits, the lab-
oratory must take corrective action and recertify the performance
C-46
-------
(Section 14.2) before proceeding with analyses. The warning and
control limits should be continuously updated as more check sample
replicates are added to the data base.
14.4 Before processing any samples, the analyst should demonstrate
through the analysis of a reagent blank that all glassware and
reagent interferences are under control. Each time a set of sam-
ples is analyzed or there is a change in reagents, a laboratory
reagent blank should be processed as a safeguard against contami-
nation.
14.5 Procedural QC - The various steps of the analytical procedure
should have quality control measures. These include but are not
limited to:
14.5.1 GC performance - See Section 7.1 for performance criteria.
14.5.2 MS performance - See Section 7.2 for performance criteria.
14.5.3 Qualitative identification - At least 10% of the PCB
identifications, as well as any questionable results,
should be confirmed by a second mass spectrometrist.
14.5.4 Quantitation - At least 10% of all manual calculations,
including peak area calculation, must be checked. After
changes in computer quantitation routes, the results
should be manually checked.
14.6 A minimum of 10% of all samples, one sample per month or one sam-
ple per matrix type, whichever is greater, must be selected at
random, sampled, and analyzed in triplicate to monitor the preci-
sion of the analysis. An RSD of ±30% or less must be achieved.
If the precision is greater than ±30%, the analyst must be re-
certified (see Section 14.2).
14.7 A minimum of 10% of all samples, one sample per month or one sam-
ple per matrix type, whichever is greater, selected at random,
must be analyzed by the standard addition technique. Two aliquots
of the sample are analyzed, one "as is" and one spiked with a suf-
ficient amount of solution CSxxx to yield approximately 100 |jg/
sample of each compound. The spiking compounds are thoroughly
incorporated by mechanical agitation. For the liquid impinger
contents, shaking for 30 sec should be sufficient. For the
Florisil, 10 min tumbling is recommended. For filters where in-
adequate incorporation may be expected, overnight equilibration
with agitation is recommended.
Note: The volume measurement of the spiking solution is critical
to the overall method precision. The analyst must exercise cau-
tion that the volume is known to ±1% or better. Where necessary,
calibration of the pipet is recommended.
C-47
-------
The samples are analyzed together and the quantitative results
calculated. The recovery of the spiked compounds (calculated by
difference) must be 80-120%. If the sample is known to contain
specific PCB isomers, these isomers may be substituted for solu-
tion CSxxx. If the concentrations of PCBs are known to be high,
the amount added should be adjusted so that the spiking level is
1.5 to 4 times the measured PCB level in the unspiked sample.
14.8 Sampling efficiency - The efficiency of PCB collection during
sampling should be monitored. This may be achieved by adding a
known amount of the 13C surrogate spiking solution (Section 6.4)
sufficient to give an analytical signal well above background to
the first impinger prior to sampling. The recovery of the four
compounds should be > 80%.
14.9 Interlaboratory comparison - Interlaboratory comparison studies
are planned. Participation requirements, level of performance,
and the identity of the coordinating laboratory will be presented
in later revisions.
14.10 It is recommended that the participating laboratory adopt addi-
tional QC practices for use with this method. The specific prac-
tices that are most productive depend upon the needs of the lab-
oratory and the nature of the samples. Field duplicates or
triplicates may be analyzed to monitor the precision of the sam-
pling technique. Whenever possible, the laboratory should per-
form analysis of standard reference materials and participate in
relevant performance evaluation studies.
15.0 Quality Assurance
Each participating laboratory must develop a quality assurance plan ac-
cording to EPA guidelines.5 The quality assurance plan must be submitted
to the Agency for approval.
16.0 Method Performance
The method performance is being evaluated. Limits of quantitation;
average intralaboratory recoveries, precision, and accuracy; and inter-
laboratory recoveries, precision, and accuracy will be presented.
17.0 Documentation and Records
Each laboratory is responsible for maintaining full records of the analy-
sis. Laboratory notebooks should be used for handwritten records. GC/MS
data must be archived on magnetic tape, disk, or a similar device. Hard
copy printouts may be kept in addition if desired. QC records should
be maintained separately from sample analysis records.
C-48
-------
The documentation must describe completely how the analysis was performed.
Any variances from the protocol must be noted and fully described. Where
the protocol lists options (e.g., sample cleanup), the option used and
specifies (solvent volumes, digestion times, etc.) must be stated.
C-49
-------
REFERENCES
1. Haile, C. L., and E. Baladi, "Methods for Determining the Polychlorinated
Biphenyl Emissions from Incineration and Capacitor and Transformer Filling
Plants," U.S. Environmental Protection Agency, (1977) EPA-600/4-73-048.
2. U.S. Environmental Protection Agency, Federal Register, 4j2(160), Thursday,
August 18, 1977.
3. Martin, R. M., "Construction Details of Isokinetic Source Sampling Equip-
ment," Environmental Protection Agency, Air Pollution Control Office
Publication No. APTD-0581.
4. Bellar, T. A., and J. J. Lichtenberg, "The Determination of Polychlorinated
Biphenyls in Transformer Fluid and Waste Oils," Prepared for U.S. Environ-
mental Protection Agency, (1981) EPA-600/4-81-045.
5. Quality Assurance Program Plan for the Office of Toxic Substances, Office
of Pesticides and Toxic Substances, U.S. Environmental Protection Agency,
Washington, D.C., October 1980.
C-50
-------
APPENDIX D
ANALYTICAL METHOD: THE ANALYSIS OF BY-PRODUCT CHLORINATED
BIPHENYLS IN INDUSTRIAL WASTEWATER
D-l
-------
THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS
IN INDUSTRIAL WASTEWATER
1.0 Scope and Application
1.1 This is a gas chromatographic/electron impact mass spectrometric
(GC/EIMS) method applicable to the determination of chlorinated
biphenyls (PCBs) in industrial wastewater. The PCBs present may
originate either as synthetic by-products or as contaminants de-
rived from commercial PCB products (e.g., Aroclors). The PCBs
may be present as single isomers or complex mixtures and may in-
clude all 209 congeners from monochlorobiphenyl through deca-
chlorobiphenyl listed in Table 1.
1.2 The detection and quantitation limits are dependent upon the vol-
ume of sample extracted the complexity of the sample matrix and
the ability of the analyst to remove interferents and properly
maintain the analytical system. The method accuracy and preci-
sion will be determined in future studies.
1.3 This method is restricted to use by or under the supervision of
analysts experienced in the use of gas chromatography/mass spec-
trometry (GC/MS) and in the interpretation of gas chromatograms
and mass spectra. Prior to sample analysis, each analyst must
demonstrate the ability to generate acceptable results with this
method by following the procedures described in Section 14.2.
1.4 The validity of the results depends on equivalent recovery of the
analyte and 13C PCBs. If the *3C PCBs are not thoroughly incor-
porated in the matrix, the method is not applicable.
1.5 During the development and testing of this method, certain analyti-
cal parameters and equipment designs were found to affect the valid-
ity of the analytical results. Proper use of the method requires
that such parameters or designs must be used as specified. These
items are identified in the text by the word "must." Anyone wish-
ing to deviate from the method in areas so identified must demon-
strate that the deviation does not affect the validity of the data.
Alternative test procedure approval must be obtained from the
Agency. An experienced analyst may make modifications to param-
eters or equipment identified by the term "recommended." Each
time such modifications are made to the method, the analyst must
repeat the procedure in Section 14.2. In this case, formal ap-
proval is not required, but the documented data from Section 14.2
must be on file as part of the overall quality assurance program.
D-2
-------
TABLE 1. NUMBERING OF PCB CONGENERS3
NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
IS
IS
17
13
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
4«
47
48
49
50
51
Structure
Mo«aeMorob
-------
2.0 Summary
2.1 The wastewater must be sampled such that the specimen collected
for analysis is representative of the whole. Statistically
designed selection of the sampling position (valve, port, outfall,
etc.) or time should be employed. The sample must be preserved to
prevent PCB loss prior to analysis. Storage at 4°C with optional
preservation at low pH is recommended.
2.2 The sample is mechanically homogenized and subsampled if necessary.
The sample is then spiked with four 13C PCB surrogates and the
surrogates incorporated by further mechanical agitation.
2.3 The surrogate-spiked sample is extracted and cleaned up at the
discretion of the analyst. Possible extraction techniques include
liquid-liquid partition and sorption onto resin columns followed
by solvent elution. Cleanup techniques may include liquid-liquid
partition, sulfuric acid cleanup, saponification, adsorption chro-
matography, gel permeation chromatography or a combination of
cleanup techniques. The sample is diluted or concentrated to a
final known volume for instrumental determination. The EPA Method
6081 and 6252 extraction and cleanup procedures may be used.
2.4 The PCB content of the sample extract is determined by capillary
(preferred) or packed column gas chromatography/electron impact
mass spectrometry (CGC/EIMS or PGC/EIMS) operated in the selected
ion monitoring (SIM), full scan, or limited mass scan (LMS) mode.
2.5 PCBs are identified by comparison of their retention time and
mass spectral intensity ratios to those in calibration standards.
2.6 PCBs are quantitated against the response factors for a mixture
of 11 PCB congeners, using the response of the 13C surrogate to
compensate for losses in workup and instrument variability.
2.7 The PCBs identified by the SIM technique may be confirmed by full
scan CGC/EIMS, retention on alternate GC columns, other mass spec-
trometric techniques, infrared spectrometry, or other techniques,
provided that the sensitivity and selectivity of the technique is
demonstrated to be comparable or superior to GC/EIMS.
2.8 The analysis time is dependent on the extent of workup employed.
The time required for instrumental analysis, excluding data re-
duction and reporting, is about 30 to 45 min.
2.9 Appropriate quality control (QC) procedures are included to assess
the performance of the analyst and estimate the quality of the
results. These QC procedures include the demonstration of labora-
tory capability: periodic analyst certification, the use of con-
trol charts, and the analysis of blanks, replicates, and standard
addition samples. A quality assurance (QA) plan must be developed
for each laboratory.
D-4
-------
2.10 While several options are available throughout this method, the
recommended procedure to be followed is:
2.10.1 The sample is collected according to a scheme which per-
mits extrapolation of the sample data to the body or con-
tainers of water being sampled.
2.10.2 The sample is preserved at low pH and at 4°C to prevent
any loss of PCBs or changes in matrix which may adversely
affect recovery.
2.10.3 The sample is-mechanically homogenized and subsampled if
necessary.
2.10.4 The sample is spiked with four 13C-PCB surrogates
(4-chlorobiphenyl; 3,3',4,4'-tetrachlorobiphenyl;
2,2',3,3',5,5',6,6'-octachlorobiphenyl; and decachloro-
biphenyl).
2.10.5 The sample is extracted.
2.10.6 The extract is cleaned up and concentrated to an appro-
priate volume.
2.10.7 An aliquot of the extract is analyzed by CGC/EIMS oper-
ated in the SIM mode. On-column injections onto a 15-m
DB-5 capillary column, programmed (for toluene solutions)
from 110° to 325°C at 10°/min after a 2 rain hold is used.
Helium at 45-cm/sec linear velocity is used as the carrier
gas.
2.10.8 PCBs are identified by retention time and mass spectral
intensities.
2.10.9 PCBs are quantitated against the response factors for a
mixture of 11 PCB congeners.
2.10.10 The total PCBs are obtained by summing the amounts for
each homolog found and the concentration is reported as
micrograms per liter.
3.0 Interferences
3.1 Method interferences may be caused by contaminants in solvents,
reagents, glassware, and other sample processing hardware, leading
to discrete artifacts and/or elevated baselines in the total ion
current profiles. All of these materials must be routinely demon-
strated to be free from interferences by the analysis of laboratory
reagent blanks as described in Section 14.4.
D-5
-------
4.0
3.1.1 Glassware must be scrupulously cleaned. All glassware
is cleaned as soon as possible after use by rinsing with
the last solvent used. This should be followed by deter-
gent washing with hot water and rinses with tap water and
reagent water. The glassware should then be drained dry
and heated in a muffle furnace at 400°C for 15 to 30 min.
Some thermally stable materials, such as PCBs, may not
be eliminated by this treatment. Solvent rinses with
acetone and pesticide quality hexane may be substituted
for the muffle furnace heating. Volumetric ware should
not be heated in a muffle furnace. After it is dry and
cool, glassware should be sealed and stored in a clean
environment to prevent any accumulation of dust or other
contaminants. It is stored inverted or capped with
aluminum foil.
3.1.2 The use of high purity reagents and solvents helps to
minimize interference problems. Purification of solvents
by distillation in all-glass systems may be required.
All solvent lots must be checked for purity prior to use.
3.2 Matrix interferences may be caused by contaminants that are coex-
tracted from the sample. The extent of matrix interferences will
vary considerably from source to source, depending upon the nature
and diversity of the sources of samples.
4.1 The toxicity or carcinogenicity of each reagent used in this
method has not been precisely defined; however, each chemical
compound should be treated as a potential health hazard. From
this viewpoint, exposure to these chemicals must be reduced to
the lowest possible level by whatever means available. The labor-
atory is responsible for maintaining a current awareness file of
OSHA regulations regarding the safe handling of the chemicals spe-
cified in this method. A reference file of material data handling
sheets should also be made available to all personnel involved in
the chemical analysis.
4.2 Polychlorinated biphenyls have been tentatively classified as known
or suspected human or mammalian carcinogens. Primary standards
of these toxic compounds should be prepared in a hood. Personnel
must wear protective equipment, including gloves and safety glasses.
Congeners highly substituted at the meta and para positions and
unsubstituted at the ortho positions are reported to be the most
toxic. Extreme caution should be taken when handling these com-
pounds neat or in concentration solution. The class includes
3,3',4,4'-tetrachlorobiphenyl (both natural abundance and isotop-
ically labeled).
D-6
-------
4.3 Diethyl ether should be monitored regularly to determine the perox-
ide content. Under no circumstances should diethyl ether be used
with a peroxide content in excess of 50 ppm as an explosion could
result. Peroxide test strips manufactured by EM Laboratories
(available from Scientific Products Company, Cat. No. P1126-8 and
other suppliers) are recommended for this test. Procedures for
removal of peroxides from diethyl ether are included in the in-
structions supplied with the peroxide test kit.
4.4 Waste disposal must be in accordance with RCRA and applicable
state rules.
5.0 Apparatus and Materials
5.1 Sampling containers - Amber glass bottles, 1-liter or other ap-
propriate volume, fitted with screw caps lined with Teflon.
Cleaned foil may be substituted for Teflon if the sample is not
corrosive. If amber bottles are not available, samples should
be protected from light using foil or a light-tight outer con-
tainer. The bottle must be washed, rinsed with acetone or methy-
lene chloride, and dried before use to minimize contamination.
5.2 Glassware - All specifications are suggestions only. Catalog
numbers are included for illustration only.
5.2.1 Volumetric flasks - Assorted sizes.
5.2.2 Pipets - Assorted sizes, Mohr delivery.
5.2.3 Micro syringes - 10.0 pi for packed column GC analysis,
1.0 pi for on-column CGC analysis.
5.2.4 Chromatographic column - Chromaflex, 400 mm long x 19 mm
ID (Kontes K-420540-9011 or equivalent).
5.2.5 Gel permeation chromatograph - GPC Autoprep 1002
(Analytical Bio Chemistry Laboratories, Inc.) or
equivalent.
5.2.6 Kuderna-Danish Evaporative Concentrator Apparatus
5.2.6.1 Concentrator tube - 10 ml, graduated (Kontes
K-570050-1025 or equivalent). Calibration must
be checked. Ground glass stopper size (S19/22
joint) is used to prevent evaporation of solvent.
5.2.6.2 Evaporative flask - 500 ml (Kontes K-57001-0500
or equivalent). Attach to concentrator tube
with springs (Kontes K-662750-0012 or equivalent),
5.2.6.3 Snyder column - Three ball macro (Kontes K503000-
0121 or equivalent).
D-7
-------
5,3 Balance - Analytical, capable of accurately weighing 0.0001 g.
5.4 Gas chromatography/mass spectrometer system.
5.4.1 Gas chromatograph - An analytical system complete with a
temperature programmable gas chromatograph and all re-
quired accessories including syringes, analytical columns,
and gases. The injection port must be designed for on-
column injection when using capillary columns or packed
columns. Other capillary injection techniques (split,
splitless, "Grob," etc.) may be used provided the per-
formance specifications stated in Section 7.1 are met.
5.4.2 Capillary GC column - A 12-20 m long x 0.25 mm ID fused
silica column with a 0.25 |Jm thick DB-5 bonded silicone
liquid phase (J&W Scientific) is recommended. Alternate
liquid phases may include OV-101, SP-2100, Apiezon L,
Dexsil 300, or other liquid phases which meet the perfor-
mance specifications stated in Section 7.1.
5.4.3 Packed GC column - A 180 cm x 0.2 cm ID glass column
packed with 3% SP-2250 on 100/120 mesh Supelcoport or
equivalent is recommended. Other liquid phases which
meet the performance specifications stated in Section 7.1
may be substituted.
5.4.4 Mass spectrometer - Must be capable of scanning from 150
to 550 Daltons every 1.5 sec or less, collecting at least
five spectra per chromatographic peak, utilizing a 70-eV
(nominal) electron energy in the electron impact ioniza-
tion mode and producing a mass spectrum which meets all
the criteria in Table 2 when 50 ng of decafluorotriphenyl
phosphine [DFTPP, bis(perfluorophenyl)phenyl phosphine]
is injected through the GC inlet. Any GC-to-MS interface
that gives acceptable calibration points at 10 ng per
injection for each PCB isomer in the calibration standard
and achieves all acceptable performance criteria (Section
10) may be used. Direct coupling of the fused silica
column to the MS is recommended. Alternatively, GC-to-
MS interfaces constructed of all glass or glass-lined
materials are recommended. Glass can be deactivated by
silanizing with dichlorodimethylsilane.
5.4.5 A computer system that allows the continuous acquisition
and storage on machine-readable media of all mass spectra
obtained throughout the duration of the chromatographic
program must be interfaced to the mass spectrometer.
The data system must have the capability of integrating
the abundances of the selected ions between specified
limits and relating integrated abundances to concentra-
tions using the calibration procedures described in this
method. The computer must have software that allows
D-8
-------
TABLE 2. DFTPP KEY IONS AND ION ABUNDANCE CRITERIA
Mass . Ion abundance criteria
197 Less than 1% of mass 198
198 100% relative abundance
199 5-9% of mass 198
275 10-30% of mass 198
365 Greater than 1% of mass 198
441 Present, but less than mass 443
442 Greater than 40% of mass 198
443 17-23% of mass 442
D-9
-------
searching any GO/MS data file for ions of a specific mass
and plotting such ion abundances versus time or scan num-
ber to yield an extracted ion current profile (EICP).
Software must also be available that allows integrating
the abundance in any EICP between specified time or scan
number limits.
6.0 Reagents
6.1 Solvents - All solvents must be pesticide residue analysis grade.
New lots should be checked for purity by concentrating an aliquot
by at least as much as is used in the procedure.
6.2 Stock standard solutions - Standards of the PCB congeners listed
in Table 3 are available from Ultra Scientific, Hope, Rhode Island;
or Analabs, North Haven, Connecticut.
6.3 Calibration standard stock solutions - Primary dilutions of each
of the individual PCBs listed in Table 3 are prepared by weighing
approximately 1-10 mg of material within 1% precision. The PCB
is then dissolved and diluted to 1.0 ml with hexane. Calculate
the concentration in mg/ml. The primary dilutions are stored at
4°C in screw-cap vials with Teflon cap liners. The meniscus is
marked on the vial wall to monitor solvent evaporation. Primary
dilutions are stable indefinitely if the seals are maintained.
The validity of primary and secondary dilutions must be monitored
on a quarterly basis by analyzing four quality control check sam-
ples (see Section 14.2).
6.4 Working calibration standards - Working calibration standards are
prepared that are similar in PCB composition and concentration to
the samples by mixing and diluting the individual standard stock
solutions. Example calibration solutions are shown in Table 3.
The mixture is diluted to volume with pesticide residue analysis
quality hexane. The concentration is calculated in ng/ml as the
individual PCBs. Dilutions are stored at 4°C in narrow-mouth,
screw-cap vials with Teflon cap liners. The meniscus is marked
on the vial wall to monitor solvent evaporation. These secondary
dilutions can be stored indefinitely if the seals are maintained.
These solutions are designated "CSxxx," where the xxx is used to
encode the nominal concentration in ng/ml.
6.5 Alternatively, certified stock solutions similar to those listed
in Table 3 may be available from a supplier, in lieu of the pro-
cedures described in Section 6.4.
6.6 DFTPP standard - A SO-ng/pl solution of DFTPP is prepared in ace-
tone or another appropriate solvent.
6.7 Surrogate standard stock solution - The four 13C-labeled PCBs
listed in Table 4 may be available from a supplier as a certified
solution. This solution may be used as received or diluted
further. These solutions are designated "SSxxx," where the xxx
is used to encode the nominal concentration in ng/ml.
D-10
-------
TABLE 3. CONCENTRATIONS OF CONGENERS IN PCS CALIBRATION STANDARDS (ng/ml)a
Homo log
1
1
2
3
4
5
6
7
8
9
10
4
1
4
8
10
Congener
no.
1
3
7
30
50
:97
143
183
202
207
209
210 (IS)
211 (RS)
212 (RS)
213 (RS)
214 (RS)
CS1000
1,040
1,000
1,040
1,040
1,520
1,740
1,920
2,600
4,640
5,060
4,240
255
104
257
407
502
CS100
104
100
104
104
152
174
192
260
464
506
424
255
104
257
407
502
CS050
52
50
52
52
76
87
96
130
232
253
212
255
104
257
407
502
CS010
10
10
10
10
15
17
19
26
46
51
42
255
104
257
407
502
a Concentrations given as examples only.
D-ll
-------
TABLE 4. COMPOSITION OF SURROGATE SPIKING SOLUTION (SS100)
CONTAINING 13C-LABELED PCBsa
Congener
no.
211
212
213
214
Compound
(I1 ,2' ,3' ,4' ,5' ,6'-13C6)4-chlorobiphenyl
(13C12)3,3' ,4,4'-tetraqhlorobiphenyl
(13C12)2,2' ,3,3' ,5,5' ,6,6'-octachlorobiphenyl
( 1 3C ! 2 ) decachlorobiphenyl
Concentration
(Mg/ml)
104
257
395
502
a Concentrations given as examples only.
D-12
-------
6.8 Internal standard solution - A solution of de-3,3' ,4,4'-tetra-
chlorobiphenyl is prepared at a nominal concentration of 1-10
rag/ml in hexane. The solution is further diluted to give a work-
ing standard.
6.9 Solution stability - The calibration standard, surrogate and
DFTPP solutions should be checked frequently for stability.
These solutions should be replaced after 6 months, or sooner if
comparison with quality control check samples indicates compound
degradation or concentration change.
6.10 Quality control check samples will be supplied by the Agency.
7.0 Calibration
7.1 The gas chromatograph must meet the minimum operating parameters
shown in Tables 5 and 6, daily. If all of the criteria are not
met, the analyst must adjust conditions and repeat the test until
all criteria are met.
7.2 The mass spectrometer must meet the minimum operating parameters
shown in Tables 2, 7, and 8, daily. If all criteria are not met,
the analyst must retune the spectrometer and repeat the test un-
til all conditions are met.
7.3 The PCB response factor (RF ) must be determined using Equation
7-1 for the analyte homologi.
A x M.
"p = £^
where RF = response factor of a given PCB isomer
A = area of the characteristic ion for the PCB congener
^ peak
M = mass of PCB congener injected (nanograms)
A. = area of the characteristic ion for the internal
18 standard peak
M. = mass of internal standard injected (nanograms)
IS
Using the same conditions as for RF , the surrogate response
factors (RF ) must be determined using Equation 7-2.
A x M.
IS S
where A = area of the characteristic ion for the surrogate peak
S
M = mass of surrogate injected (nanograms)
S
Other items are the same as defined in Equation 7-1.
D-13
-------
TABLE 5. OPERATING PARAMETERS FOR CAPILLARY COLUMN GAS CHROMATOGRAPHIC SYSTEM
Parameter
Recommended
Tolerance
Gas chromatograph
Column
Liquid phase
Liquid phase thickness
Carrier gas
Carrier gas velocity
Injector
Injector temperature
Injection volume
Initial column temperature
Column temperature program
Finnigan 9610
15 m x 0.255 mm ID
Fused silica
DB-5 (J&W)
0.25 \M
Helium
/c , b
45 cm/sec
p
On-column (J&W)
Optimum performance
1.0 |jlc
70°C (2 min)d
70°-325°C at 10°C/min£
Other
Other
Other nonpolar
or semipolar
< 1 |Jm
Hydrogen
Optimum performance
Other
Optimum performance
Other
Other
Other
Separator
Transfer line temperature
Tailing factor
Peak width
None
280°C
0.7-1.5
7-10 sec
Glass jet or othe
Optimum*
0.4-3
< 15 sec
a Substitutions permitted with any common apparatus or technique provided
performance criteria are met.
b Measured by injection of air or methane at 270°C oven temperature.
c For on-column injection, manufacturer's instructions should be followed
regarding injection technique.
d With on-column injection, initial temperature equals boiling point of the
solvent; in this instance, hexane.
e C12C11Q elutes at 270°C. Programming above this temperature ensures a
clean column and lower background on subsequent runs.
f Fused silica columns may be routed directly into the ion source to pre-
vent separator discrimination and losses.
g High enough to elute all PCBs, but not high enough to degrade the column
if routed through the transfer line.
h Tailing factor is width of front half of peak at 10% height divided by width
of back half of peak at 10% height for single PCB congeners in solution CSxxx.
i Peak width at 10% height for a single PCB congener is CSxxx.
D-14
-------
TABLE 6. OPERATING PARAMETERS FOR PACKED COLUMN GAS CHROMATOGRAPHY SYSTEM
Parameter
Recommended
Tolerances
Gas chromatograph
Column
Finnigan 9610
180 cm x 0.2 cm ID
Other3
Other
Column packing
Carrier gas
Carrier gas flow.rate
Injector
Injector temperature
Injection volume
Initial column temperature
Column temperature program
Separator
Transfer line temperature
Tailing factor
Peak width
glass
3% SP-2250 on 100/
120 mesh Supelcoport
Helium
30 ml/min
On-column
250°C
1.0 |jl
150°C, 4 min
150°C-260° at 8°/min
Glass jet
280°C
0.7-1.5
10-20 sec
Other nonpolar
or semipolar
Hydrogen
Optimum performance
Optimum
^ 5 |Jl
Other
Other
Other
Optimum
0.4-3
< 30 sec
a Substitutions permitted if performance criteria are met.
b High enough to elute all PCBs.
c Tailing factor is width of front half of peak at 10% height divided by
width of back half of peak at 10% height for single PCB congeners in solu-
tion CSxxx.
d Peak width at 10% height for a single PCB congener in CSxxx.
D-15
-------
TABLE 7. OPERATING PARAMETERS FOR QUADRUPOLE MASS SPECTROMETER SYSTEM
Parameter Recommended Tolerance
Mass spectrometer
Data system
Scan range
Scan time
Resolution
Ion source temperature
Electron energy
Trap current
Multiplier voltage
Preamplifier sensitivity
Finnigan 4023
Incos 2400
95-550
1 sec
Unit
280°C
70 eV
0.2 mA
-1,600 V
10"6 A/V
Other3
Other
Other
Otherb
Optimum performance
200°-300°C
Optimum performance
Optimum performance
Optimum performance
Set for desired
working range
a Substitutions permitted if performance criteria are met.
b Greater than five data points over a GC peak is a minimum.
c Filaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no separator is used.
D-16
-------
TABLE 8. OPERATING PARAMETERS FOR MAGNETIC SECTOR MASS SPECTROMETER SYSTEM
Parameter
Mass spectrometer
Data system
Scan range
Scan mode
Cycle time
Resolution
Ion source temperature
Electron energy
Emission current
Filament current
Multiplier
Recommended
Finnigan MAT 31 1A
Incos 2400
98-550
Exponential
1.2 sec
1,000
280°C
70 eV
1-2 mA
Optimum
-1,600 V
Tolerance
Other3
Other
Other
Other
Otherb
> 500
250°-300°C
70 eV
Optimum
Optimum
Optimum
a Substitutions permitted if performance criteria are met.
b Greater than five data points over a GC peak is a minimum.
c Filaments should be shut off "during solvent elution to improve instrument
stability and prolong filament life, especially if no separator is used.
D-17
-------
If specific congeners are known to be present and if standards
are available, selected RF values may be employed. For general
samples, solutions CSxxx and SSxxx or a mixture (Tables 3 and 4),
with a similar level of internal standard (de-3,31,4,4'-tetra-
chlorobiphenyl) added, may be used as the response factor solution.
The PCB-surrogate pairs to be used in the RF calculation are listed
in Table 9.
Generally, only the primary ions of both the analyte and surrogate
are used to determine the RF values. If alternate ions are to be
used in the quantitation, the RF must be determined using that
characteristic ion.
The RF value must be determined in a manner to assure ±20% accu-
racy and precision. For instruments with good day-to-day preci-
sion, a running mean (RF) based on seven values determined once
each day may be appropriate. Other options include, but are not
limited to, triplicate determinations of a single concentration
spaced throughout a day or determination of the RF at three dif-
ferent levels to establish a working curve.
If replicate RF values differ by greater than ±10% RSD, the system
performance should be monitored closely. If the RSD is greater
than ±20%, the data set must be considered invalid and the RF re-
determined before further analyses are done.
7.4 If the GC/EIMS system has not been demonstrated to yield a linear
response or if the analyte concentrations are more than two orders
of magnitude different from those in the RF solution, a calibration
curve must be prepared. If the analyte and RF solution concentra-
tions differ by more than one order of magnitude, a calibration
curve should be prepared. A calibration curve should be estab-
lished with triplicate determinations at three or more concentra-
tions bracketing the_ analyte levels.
7.5 The relative retention time (RRT) windows for the 10 homologs and
surrogates must be determined. If all congeners are not available,
a mixture of available congeners or an Aroclor mixture (e.g.,
1016/1254/1260) may be used to estimate the windows. The windows
must be set wider than observed if all isomers are not determined.
Typical RRT windows for one column are listed in Table 10. The
windows may differ substantially if other GC parameters are used.
8.0 Sample Collection, Handling, and Preservation
8.1 Amber glass sample containers should have Teflon-lined screw caps.
With noncorrosive samples, methylene chloride-washed aluminum foil
liners may be substituted. The volume is determined by the amount
of sample to be collected but will usually be 1 liter or 1 qt.
The sample size is dependent on the anticipated PCB levels and
difficulty of the subsequent extraction/cleanup steps.
D-18
-------
TABLE 9. PAIRINGS OF ANALYTE, CALIBRATION, AND SURROGATE COMPOUNDS
vO
Analyte
Congener
no.
1
2,3
4-15
16-39
40-81
82-127
128-169
170-193
194-205
206-208
209
Compound
Calibration standard
Congener
no.
2"Ci2HgCl 1
3- and 4-C12H9Cl 3
C12H8C12
C12H7C13
C12H6C14
C12H5C15
C12H3C17
C12H2C18
C12HC19
C12C110
7
30
50
97
143
183
202
207
209
2
4
2
2
2
2
2
2
2
2
C
,4
,4,
,2'
,2'
,2'
,2'
,2'
,2'
12C
6
,4
,3
,3
,3
,3
,3
:ii
Compound
,6
i
,4
i
,3
,3
0
4,5
,5,6'
4, 4', 5', 6
',5,5', 6, 6'
',4, 4', 5, 6, 6'
Surrogate
Congener
no.
211
211
211
212
212
212
212
213
213
213
214
Compound
13C6-4
13C6-4
13C6-4
13Ci2-3,3'
13C12-3,3'
13C12-3,3'
13C12-3,3'
13Ci2-2,2'
13Ci2-2,2'
13C12-2,2'
13C12Clio
,4,4'
,4,4'
,4,4'
,4,4'
,3,3'
0 01
,J,J
0 01
,0,0
,5, 5', 6, 6'
,5, 5', 6, 6'
,5, 5', 6, 6'
a Ballschmiter numbering system, see Table 1.
-------
TABLE 10. RELATIVE RETENTION TIME (RRT) RANGES OF PCB HOMOLOGS
VERSUS d6-3,3'.4,4'-TETRACHLOROBIPHENYL
PCB
homolog
Monochloro
Dichloro
Trichloro
Tetrachloro
Pentachloro
Hexachloro
Heptachloro
Octachloro
Nonachloro
Decachloro
No. of
isomers
measured
3
10
9
16
12
13
4
6
3
1
Calibration solution
Observed range
of RRTsa
0.40-0.50
0.52-0.69
0.62-0.79
0.72-1.01
0.82-1.08
0.93-1.20
1.09-1.30
1.19-1.36
1.31-1.42
1.44-1.45
Congener
no.
1
3
7
30
50
97
143
183
202
207
209
Observed
RRT3
0.43
0.50
0.58
0.65
0.75
0.98
1.05
1.15
1.19
1.33
1.44
Projected
range of
RRTsD
0.35-0.55
0.35-0.80
0.35-1.10
0.55-1.05
0.80-1.10
0.90-1.25
1.05-1.35
1.10-1.50
1.25-1.50
1.35-1.50
a The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were
measured versus 3,3',4,4'-tetrachlorobiphenyl-de (internal standard) using
a 15-m J&W DB-5 fused silica 'column with a temperature program of 110°C
for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an on-
column injector. A Finnigan 4023 Incos quadrupole mass spectrometer oper-
ating with a scan range of 95-550 daltons was used to detect each PCB
congener.
b The projected relative retention windows account for overlap of eluting
homologs and take into consideration differences in operating systems and
lack of all possible 209 PCB congeners.
D-20
-------
8.2 Sample bottle preparation
8.2.1 All sample bottles and caps should be washed in detergent
solution, rinsed with tap water and then with distilled
water. The bottles and caps are allowed to drain dry in
a contaminant-free area. Then the caps are rinsed with
pesticide grade hexane and allow to air dry.
8.2.2 Sample bottles are heated to 400°C for 15 to 20 min or
rinsed with pesticide grade acetone or hexane and allowed
to air dry.
8.2.3 The clean bottles are stored inverted or sealed until use.
8.3 Sample collection
8.3.1 The primary consideration in sample collection is that
the sample collected be representative of the whole.
Therefore, sampling plans or protocols for each individ-
ual producer's situation will have to be developed. The
recommendations presented here describe general situa-
tions. The number of replicates and sampling frequency
also must be planned prior to sampling.
8.3.2 If possible, mix the source thoroughly before collecting
the sample. If mixing is impractical, the sample should
be collected from a representative area of the source.
If the liquid is flowing through an enclosed system, sam-
pling through a valve should be randomly timed.
8.3.3 Fill the bottle with water, add preservative (Section
8.4), cap tightly, and shake well. To prevent the caps
from working loose during storage tape the caps on with A
a water-insoluble tape.
8.4 Sample preservation - Samples should be stored at 4°C. Since
there is a possibility of microbial degradation, addition of H2S04
during collection to a pH < 2 is recommended. A test strip is
used to monitor the pH. Storage times in excess of 4 weeks are
not recommended.
If residual chlorine is present in the sample, it should be
quenched with sodium thiosulfate. EPA Methods 330.4 and 330.5
may be used to measure the residual chlorine.3 Field test kits
are available for this purpose.
9.0 Sample Preparation
9.1 Sample homogenization and subsampling - The sample is homogenized
by shaking, blending, or other appropriate mechanical technique,
if necessary. If the density of the sample is not between 0.9
D-21
-------
and 1.1, the density should be determined and reported. Consider-
ation should be given to treating the sample as a product waste
(see separate protocol).
Note: The precision of the mass determination at this step will
be reflected in the overall method precision. Therefore,
an analytical balance must be used to assure that the
weight is accurate to ±1% or better.
9.2 Surrogate addition - An appropriate volume of surrogate solution
SSxxx is pipetted into the sample. The final concentration of the
surrogates must be in the working range of the calibration and
well above the matrix background.
Note: The volume measurement of the spiking solution is criti-
cal to the overall method precision. The analyst must
exercise caution that the volume is known to ±1% or
better. Where necessary, calibration of the pipet is
recommended.
9.3 Sample preparation (extraction/cleanup) - After addition of the
surrogates, the sample is further treated at the discretion of
the analyst, provided that the GC/EIMS response of the four sur-
rogates meets the criteria listed in Section 7.0. The literature
pertaining to these techniques has been reviewed.4 Several pos-
sible techniques are presented below for guidance only. The ap-
plicability of any of these techniques to a specific sample matrix
must be determined by the precision and accuracy of the C PCB
surrogate recoveries, as discussed in Section 14.2.
9.3.1 Extraction - The entire sample must be transferred to the
extraction vessel with PCB-free water washing, if neces-
sary, to transfer all solids. The container is then
rinsed with the extraction solvent to recovery any PCBs
adhering to' the container wall. The solvent rinses are
combined with the extracts from below. Measure the sam-
ple volume to the nearest 0.5%.
9.3.1.1 Liquid-liquid extraction - The solvent, number
of extractions, solvent-to-sample ratio, and
other parameters are chosen at the analyst's
discretion. A suggested extraction from water
is presented in EPA Methods 6081 and 625.2
9.3.1.2 Sorbent column extraction - PCBs may be isolated
from water onto sorbent columns, although these
techniques are not as widely used or thoroughly
validated as liquid-liquid extraction. The
selection of sorbent (XAD, Porapak, carbon-
polyurethane foam, etc.) will depend on the
nature of the matrix. The available methods
have been reviewed.4
D-22
-------
9.3.2 Cleanup - Several tested cleanup techniques are described
below. All but the base cleanup (9.3.2.8) were previously
validated for PCBs in transformer fluids.5 Depending
upon the complexity of the sample, one or more of the
techniques may be required to fractionate the PCBs from
interferences. For most cleanups a concentrated (1-5
ml) extract should be used.
9.3.2.1 Acid cleanup
9.3.2.1.1 Place 5 ml of concentrated sulfuric
acid into a 40-ml narrow-mouth screw-
cap bottle. Add the sample extract.
Seal the bottle with a Teflon-lined
screw cap and shake for 1 min.
9.3.2.1.2 Allow the phases to separate, trans-
fer the sample (upper phase) with
three rinses of 1-2 ml solvent to a
clean container and concentrate to
an appropriate volume.
9.3.2.1.3 Analyze as described in Section 10.0.
9.3.2.1.4 If the sample is highly contaminated,
a second or third acid cleanup may
be employed.
9.3.2.2 Florisil column cleanup
9.3.2.2.1 Variations among batches of Florisil
(PR grade or equivalent) may affect
the elution volume of the various
PCBs. For this reason, the volume
of solvent required to completely
elute all of the PCBs must be veri-
fied by the analyst. The weight of
Florisil can then be adjusted accor-
dingly.
9.3.2.2.2 Place a 20-g charge of Florisil,
activated overnight at 130°C, into a
Chromaflex column. Settle the Flor-
isil by tapping the column. Add
about 1 cm of anhydrous sodium sul-
fate to the top of the Florisil.
Pre-elute the column with 70-80 ml
of hexane. Just before the exposure
of the sodium sulfate layer to air,
stop the flow. Discard the eluate.
D-23
-------
9.3.2.2.3 Add the sample extract to the column.
9.3.2.2.4 Carefully wash down the inner wall
of the column with 5 ml of the hexane.
9.3.2.2.5 Add 220 ml of hexane to the column.
9.3.2.2.6 Discard the first 25 ml.
9.3.2.2.7 Collect 200 ml of hexane eluate in a
Kuderna-Danish flask. All of the
PCBs should be in this fraction.
Concentrate to an appropriate volume.
9.3.2.2.8 Analyze the sample as described in
Section 10.0.
9.3.2.3 Alumina column cleanup
9.3.2.3.1 Adjust the activity of the alumina
(Fisher A540 or equivalent) by heat-
ing to 200°C for 2 to 4 hr. When
cool, add 3% water (wt:wt) and mix
until uniform. Store in a tightly
sealed bottle. Allow the deactivated
alumina to equilibrate at least 1/2
hr before use. Reactivate weekly.
9.3.2.3.2 Variations between batches of alumina
may affect the elution volume of the
various PCBs. For this reason, the
volume of solvent required to com-
pletely elute all of the PCBs must
be verified by the analyst. The
weight of alumina can then be ad-
justed accordingly.
9.3.2.3.3 Place a 50-g charge of alumina into
a Chromaflex column. Settle the alu-
mina by tapping. Add about 1 cm of
anhydrous sodium sulfate. Pre-elute
the column with 70-80 ml of hexane.
Just before exposure of the sodium
sulfate layer to air, stop the flow.
Discard the eluate.
9.3.2.3.4 Add the sample extract to the column.
9.3.2.3.5 Carefully wash down the inner wall
of the column with 5 ml volume of
hexane.
D-24
-------
9.3.2.3.6 Add 295 ml of hexane to the column.
9.3.2.3.7 Discard the first 50 ml.
9.3.2.3.8 Collect 250 ml of the hexane in a
Kuderna-Danish flask. All of the
PCBs should be in this fraction.
Concentrate to an appropriate volume.
9.3.2.3.9 Analyze the sample as described in
Section 10.0.
9.3.2.4 Silica gel column cleanup
9.3.2.4.1 Activate silica gel (Davison grade
950 or equivalent) at 135°C overnight.
9.3.2.4.2 Variations between batches of silica
gel may affect the elution volume of
the various PCBs. For this reason,
the volume of solvent required to
completely elute all of the PCBs must
be verified by the analyst. The
weight of silica gel can then be ad-
justed accordingly.
9.3.2.4.3 Place a 25-g charge of activated
silica gel into a Chromaflex column.
Settle the silica gel by tapping the
column. Add about 1 cm of anhydrous
sodium sulfate to the top of the
silica gel.
9.3.2.4.4 Pre-elute the column with 70-80 ml
of hexane. Discard the eluate. Just
before exposing the sodium sulfate
layer to air, stop the flow.
9.3.2.4.5 Add the sample extract to the column.
9.3.2.4.6 Wash down the inner wall of the column
with 5 ml of hexane.
9.3.2.4.7 Elute the PCBs with 195 ml of 10%
diethyl ether in hexane (v:v).
9.3.2.4.8 Collect 200 ml of the eluate in a
Kuderna-Danish flask. All of the
PCBs should be in this fraction.
Concentrate to an appropriate volume.
D-25
-------
9.3.2.4.9 Analyze the sample according to Sec-
tion 10.0.
9.3.2.5 Gel permeation cleanup
9.3.2.5.1 Set up and calibrate the gel perme-
ation chromatograph with an SX-3
column according to the Autoprep in-
struction manual. Use 15% methylene
chloride in cyclohexane (v:v) as the
mobile phase.
9.3.2.5.2 Inject 5.0 ml of the sample extract
into the instrument. Collect the
fraction containing the PCBs (see
Autoprep operator's manual) in a
Kuderna-Danish flask equipped with
a 10-ml ampul.
9.3.2.5.3 Concentrate the PCB fraction to an
appropriate volume.
9.3.2.5.4 Analyze as described in Section 10.0.
9.3.2.6 Acetonitrile partition
9.3.2.6.1 Place the sample extract into a 125-ml
separatory funnel with enough hexane
to bring the final volume to 15 ml.
Extract the sample four times by shak-
ing vigorously for 1 min with 30-ml
portions of hexane-saturated acetoni-
trile.
9.3.2.6.2 Combine and transfer the acetonitrile
phases to a 1-liter separatory funnel
and add 650 ml of distilled water
and 40 ml of saturated sodium chloride
solution. Mix thoroughly for about 30
sec. Extract with two 100-ml por-
tions of hexane by vigorously shaking
about 15 sec.
9.3.2.6.3 Combine the hexane extracts in a
1-liter separatory funnel and wash
with two 100-ml portions of distilled
water. Discard the water layer and
pour the hexane layer through a 8-10
cm anhydrous sodium sulfate column
into a 500-ml Kuderna-Danish flask
equipped with a 10-ml ampul. Rinse
the separatory funnel and column with
three 10-ml portions of hexane.
D-26
-------
9.3.2.6.4 Concentrate the extracts to an
appropriate volume.
9.3.2.6.5 Analyze as described in Section 10.0.
9.3.2.7 Florisil slurry cleanup
9.3.2.7.1 Place the sample extract into a 20-ml
narrow-mouth screw-cap container.
Add 0.25 g of Florisil (PR grade or
equivalent). Seal with a Teflon-lined
screw cap and shake for 1 min.
9.3.2.7.2 Allow the Florisil to settle; then
decant the treated solution into a
second container with rinsing. Con-
centrate the sample to an appropriate
volume. Analyze as described in Sec-
tion 10.0.
9.3.2.8 Base cleanup6
9.3.2.8.1 Quantitatively transfer the concen-
trated extract to a 125-ml extraction
flask with the aid of several small
portions of solvent.
9.3.2.8.2 Evaporate the extract just to dry-
ness with a gentle stream of dry
filtered nitrogen, and add 25 ml of
2.5% alcoholic KOH.
9.3-2.8.3 Add a boiling chip, put a water con-
denser in place, and allow the solu-
tion to reflux on a hot plate for 45
min.
9.8.2.8.4 After cooling, transfer the solution
to a 250-ml separatory funnel with
25 ml of distilled water.
9.3.2.8.5 Rinse the extraction flask with 25
ml of hexane and add it to the
separatory funnel.
9.3.2.8.6 Stopper the separatory funnel and
shake vigorously for at least 1 min.
Allow the layers to separate and
transfer the lower aqueous phase to
a second separatory funnel.
D-27
-------
9.3.2.8.7 Extract the saponification solution
with a second 25-ml portion of hexane.
After the layers have separated, add
the first hexane extract to the sec-
ond separatory funnel and transfer
the aqueous alcohol layer to the
original separatory funnel.
9.3.2.8.8 Repeat the extraction with a third
25-ml portion of hexane. Discard
the saponification solution, and com-
bine the hexane extracts.
9.3.2.8.9 Concentrate the hexane layer to an
appropriate volume and analyze ac-
cording to Section 10.0.
10.0 Gas Chromatographic/Electron Impact Mass Spectrometric Determination
10.1 Internal standard addition - An appropriate volume of the internal
standard solution is pipetted into the sample. The final concen-
tration of the internal standard must be in the working range of
the calibration and well above the matrix background. The inter-
nal standard is thoroughly incorporated by mechanical agitation.
Note: The volumetric measurement of the internal standard solu-
tion is critical to the overall method precision. The analyst
must exercise caution that the volume is known to be ±1% or better.
Where necessary, calibration of the pipet is recommended.
10.2 Tables 2, and 5 through 8 summarize the recommended operating con-
ditions for analysis. Figure 1 presents an example of a chromato-
gram.
10.3 While the highest available chromatographic resolution is not a
necessary objective of this protocol, good chromatographic per-
formance is recommended. With the high resolution of CGC, the
probability that the chromatographic peaks consist of single com-
pounds is higher than with PGC. Thus, qualitative and quantita-
tive data reduction should be more reliable.
10.4 After performance of the system has been certified for the day
and all instrument conditions set according to Tables 2, and 5
through 8, inject an aliquot of the sample onto the GC column.
If the response for any ion, including surrogates and internal
standards, exceeds the working range of the system, dilute the
sample and reanalyze. If the responses of surrogates, analyte,
or internal standard are below the working range, recheck the
system performance. If necessary, concentrate the sample and
reanalyze.
D-28
-------
108.On
RIG
(V)
rH
C
0)
x;
p.
•H
XI
o
)_l
o
rH
X!
o
o
G
o
a
1
CN
G10
1
523 575 1 1
033
10:0;)
vO
CJ
CO
rH
rH rH
C C
01 01
X! XI
P. P.
iH -H
XI XI
0 0
r4 t-l
O O
rH rH
x; x:
o o
o o
c c
0 0
1 1
•o
4-J l^""K
CO 01
4-1
rH CO
cd oo
c o
01 M
4-1 3
C to
vO CO
•O rH
1 1
rH rH
t^t r^»
C C
Ol Ol
XI XI
ex p.
•H -iH
Xi Xi
o o
r^lfVV^
o
1 — 1
XI
o
cd
J^
4-1
Ol
H
1
rH -4T
rS •*
c -*
0) -
x: -
(X CO
•iH
Xi co
0
o
rH
XI
CJ
cd
4-1
c
01
i
•>
"^
ro
CN" I
-J
rH
XI
O
Cd rH
(-1 >i
4J C
01 01
H XI
1 CX
•H
ST XI
« O
-3- l-l
- O
- rH
ro XI
• O
en cd
X
01
32
I
—
vD
"
"1
-*
CO
CN
CN
1101
L
iiwe
IGHU
rH
B
O)
XI
cx
•H
O
o
rH
XI
O
cd
4-1
p.
01
33
vO
V.
•.
LO
•»
-
^
~*
n
CM K
CN"
•
1178
1
20-
m
rH
^
rH C
>-, 01
C XI
0) CX
X! -H rH
P. XI >.
•H O C
XI MO)
o ox;
t-l rH P.
o x: -H
rH O XI
xi cd o
U C t-i
cd o o
4-1 a rH
u i x:
O 0
I vo cd
o
vO vo Ol
n f. Q
VO 1A
•t n
f ~
if) -^
* •*
ir> ^f
«t n
•. •.
ro ro
A *
ro ro
•• •.
CN CN
CN CN
13IG
JL
' . 112592.
CN
rH
0
ro •
rH
|
rH
^
C
Ol
x:
P-. *-\
•H Ol
n ^j
O cd
t-i 60
0 O
*~H VH
_<^ ^
0 3
cd 10
O — '
Ol
Q
•
0 l'i'i:9 SCA'l
Ct) 23:20 Tllh:
Figure 1. Capillary gas chromatography/electron impact ionization mass spectrometry (CGC/EIMS)
chromatogram or the calibration standard solution required for quantitation of PCBs by homolog.
This chromatogram includes PCBs representative of each homolog, three carbon-13 labeled surrogates,
and the deuterated internal standard. The concentration of all components and the CGC/EIMS
parameters are presented in Tables 3, 4, 5, and 7.
-------
10.5 Record all data on a digital storage device (magnetic disk, tape,
etc.) for qualitative and quantitative data reduction as discussed
Vx /-s 1 ^\r.ir
below.
11.0 Qualitative Identification
11.1 Selected ion monitoring (SIM) or limited mass scan (IMS) data -
The identification of a compound as a given PCB homolog requires
that two criteria be met:
11.1.1 (1) The peak must elute within the retention time window
set for that homolog (Section 7.5); and (2) the ratio of
two ions obtained by SIM (Table 11) or by IMS (Table 12)
must match the natural ratio within ±20%. The analyst
must search the higher mass windows, in particular M+70,
to prevent misidentification of a PCB fragment ion clus-
ter as the parent.
11.1.2 If one or the other of these criteria is not met, inter-
ferences may have affected the results and a reanalysis
using full scan EIMS conditions is recommended.
11.2 Full scan data
11.2.1 The peak must elute within the retention time windows
set for that homolog (as described in Section 7.5).
11.2.2 The unknown spectrum must match that of an authentic PCB.
The intensity of the three largest ions in the molecular
cluster (two largest for monochlorobiphenyls) must match
the natural ratio within ±20%. Fragment clusters with
proper intensity ratios must also be present.
11.2.3 Alternatively, a spectral search may be used to automat-
ically reduce the data. The criteria for acceptable
identification include a high index of similarity. For
the Incos 2300, a fit of 750 or greater must be obtained.
11.3 Disputes in interpretation - Where there is reasonable doubt as
to the identity of a peak as a PCB, the analyst must either iden-
tify the peak as a PCB or proceed to a confirmational analysis
(see Section 13.0).
12.0 Quantitative Data Reduction
12.1 Once a chromatographic peak has been identified as a PCB, the com-
pound is quantitated based either on the integrated abundance of
the SIM data or EICP for the primary characteristic ion in Tables
11 and 12. If interferences are observed for the primary ion,
D-30
-------
TABLE 11. CHARACTERISTIC SIM IONS FOR PCBs
Homolog
Ci^HgCl
Cl2HsCl2
C^HyCla
C12H6C14
C12H5Cl5
^12^4^16
C12H3Cl7
C12H2C18
Ci2HCl9
CiaCliO
Primary
188 (100)
222 (100)
256 (100)
292 (100)
326 (100)
360 (100)
394 (100)
430 (100)
464 (100)
498 (100)
Ion (relative intensity)
Secondary
190 (33)
224 (66)
258 (99)
290 (76)
328 (66)
362 (82)
396 (98)
432 (66)
466 (76)
500 (87)
Tertiary
-
226 (11)
260 (33)
294 (49)
324 (61)
364 (36)
398 (54)
428 (87)
462. (76)
496 (68)
Source: Rote, J. W., and W. J. Morris, "Use of Isotopic Abundance Ratios in
Identification of Polychlorinated Biphenyls by Mass Spectroraetry,"
J. Assoc. Offic. Anal. Chem., 56(1), 188-199 (1973).
D-31
-------
TABLE 12. LIMITED MASS SCANNING (IMS) RANGES FOR PCBs
Compound
C^HgCli
C12H8C12
C12H7C13
C12H6C13
C^Hsds
C12H4Cl6
C12H3Cl7
C12H2C18
Cl2HClg
Cl2cll6
C12D6C14
^Cg^CgHgCl
13C12H6C14
13C12H2C18
"C12C110
Mass range (m/z)
186-190
220-226
254-260
288-294
322-328
356-364
386-400
426-434
460-468
494-504
294-300
192-1.96
300-306
438-446
506-516
a Adapted from Tindall, G. W., and P. E. Wininger, "Gas Chromatography-Mass
Spectrometry Method for Identifying and Determining Polychlorinated Bi-
phenyls," J. Chromatogr., 196, 109-119 (1980).
D-32
-------
use the secondary and then tertiary ion for quantitation. If in-
terferences in the parent cluster prevent quantitation, an ion
from a fragment cluster (e.g., M-70) may be used. Whichever ion
is used, the RF must be determined using that ion. The same cri-
teria should be applied to the surrogate compounds (Table 13).
12.2 Using the appropriate analyte-internal standard pair and response
factor (RF ) as determined in Section 7.3, calculate the concen-
tration of*each peak using Equation 12-1.
A Mig Vg
Concentration (pg/g) = •£- ' gjr- • ^ • ^ Eq. 12-1
is p e i
where A = area of the characteristic ion for the analyte PCB
p peak
A. = area of the characteristic ion for the internal
standard peak
RF = response factor of a given PCB congener
M. = mass of internal standard injected (micrograms)
IS
M = mass of sample extracted (grams)
V. = volume injected (microliters)
V = volume of sample extract (microliters)
12.3 If a peak appears to contain non-PCB interferences which cannot
be circumvented by a secondary or tertiary ion, either:
12.3.1 Reanalyze the sample on a different column which sepa-
rates the PCB and interferents;
12.3.2 Perform additional chemical cleanup (Section 9) and then
reanalyze the sample; or
12.3.3 Quantitate the entire peak as PCB.
12.4 Calculate the recovery of the four 13C surrogates using the ap-
propriate surrogate-internal standard pair and response factor
(RF. ) as determined in Section 7.4 using Equation 12-2.
1S
A M.
Recovery (%) = -f- • ^- - —^ - 100 Eq. 12-2
is s s
where A = area of the characteristic ion for the surrogate peak
S
A. = area of the characteristic ion for the internal standard
is ,
peak
D-33
-------
TABLE 13. CHARACTERISTIC IONS FOR 13C-LABELED PCB SURROGATES
Ion (relative intensity)
Specific compound Primary Secondary Tertiary
i3C612C6H9Cl 194 (100) 196 (33)
13C12H6C14 304 (100) 306 (49) 302 (78)
13C12H2C18 442 (100) 444 (65) 440 (89)
13C12C110 510 (100) 512 (87) 514 (50)
D-34
-------
RF = response factor for the surrogate compound with respect
s to the internal standard (Equation 7-2)
M. = mass of internal standard injected (nanograms)
IS
M = mass of surrogate, assuming 100% recovery (nanograms)
S
12.5 Correct the concentration of each peak using Equation 12-3. This
is the final reportable concentration.
Corrected concentration (Mg/g) = C°nRecovery°(%^/S * 10° Eq* 12'3
12.6 Sum all of the peaks for each homolog, and then sum those to yield
the total PCB concentration in the sample. Report all numbers in
[Jg/g. The reporting form in Table 14 may be used. If an alter-
nate reporting format (e.g., concentration per peak) is desired,
a different report form may be used. The uncorrected concentra-
tions, percent recovery, and corrected recovery are to be reported.
12.7 Round off all numbers reported to two significant figures.
13.0 Confirmation
If there is reason to question the qualitative identification (Section
11.0), the analyst may choose to confirm that a peak is not a PCB. Any
technique may be chosen provided that it is validated as having equiva-
lent or superior selectivity and sensitivity to GC/EIMS. Some candidate
techniques include alternate GC columns (with EIMS detection), GC/CIMS,
GC/NCIMS, high resolution EIMS, and MS/MS techniques. Each laboratory
must validate confirmation techniques to show equivalent or superior
selectivity between PCBs and interferences and sensitivity (limit of
quantitation, LOQ).
If a peak is confirmed as being a non-PCB, it may be deleted from the
calculation (Section 12). If a peak is confirmed as containing both
PCB and non-PCB components, it must be quantitated according to Section
12.3.
14.0 Quality Control
14.1 Each laboratory that uses this method must operate a formal qual-
ity control (QC) program. The minimum requirements of this pro-
gram consist of an initial demonstration of laboratory capability
and the analysis of spiked samples as a continuing check on per-
formance. The laboratory must maintain performance records to
define the quality of data that are generated. After a date spe-
cified by the Agency, ongoing performance checks should be com-
pared with established performance criteria to determine if the
results of analyses are within accuracy and precision limits ex-
pected of the method.
D-35
-------
TABLE 14. ANALYSIS REPORT
INCIDENTAL PCBs IN WASTEWATER
Sample No.
Sample Matrix
Sample Source
Notebook No. or File Location
Volume Extracted liter
Extraction/Cleanup Procedure
Int. Std. Mass Added (|Jg) (Circle one) Ratio Intensity
4-Cl(d6) 298 300 100/49
Surrogates Mass Added (|jg) (Circle one) Ratio Intensity % Recovery
1-C1 194 196 100/33
4-C1 304 306 100/49
8-C1 442 444 100/65
10-C1 510 512 100/87
(continued)
D-36
-------
TABLE 14 (continued)
Qualitative Quantitative
Analyte 1° 2° Il°
1-C1 188 190
2-C1 222 224
3-C1 256 258
4-C1 292 290
5-C1 326 328
6-C1 360 362
7-C1 394 396
8-C1 430 432
9-C1 464 466
10-C1 498 500
Total
Reported by:
Name
Signature/Date
Organization
Uncorr Corr
Ion Cone. Cone
12° Ratio Theoretical OK? Used RF (Mg/£) (Mg/£)
100/33
100/66
100/99
100/76
100/66
100/82
100/98
100/66
100/76
100/87
MS/£ M8/£
Uncorr. Corr.
Internal Audit: EPA Audit:
Name Name
Signature/Date Signature/Date
Organization Organization
D-37
-------
14.2 The analysts must certify that the precision and accuracy of the
analytical results are acceptable by:
14.2.1 The absolute precision of surrogate recovery, measured
as the RSD of the integrated EIMS area (A ) for a set
of samples, must be ±10%.
14.2.2 The mean recovery (R ) of at least four replicates of a
QC check sample to be supplied by the Agency must meet
Agency-specified accuracy and precision criteria. This
forms the initial data base for establishing control
limits (see Section 14.3 below).
14.3 Control limits - The laboratory must establish control limits
using the following equations:
Upper control limit (UCL) = RC + 3 RSDc
Upper warning limit (UWL) = R +2 RSD
Lower warning limit (LWL) = R - 2 RSD
Lower control limit (LCL) = R - 3 RSD
These may be plotted on control charts. If an analysis of a
check sample falls outside the warning limits, the analyst should
be alerted that potential problems may need correction. If the
results for a check sample fall outside the control limits, the
laboratory must take corrective action and recertify the perfor-
mance (Section 14.2) before proceeding with analyses. The warn-
ing and control limits should be continuously updated as more
check sample replicates are added to the data base.
14.4 Before processing any samples, the analyst should demonstrate
through the analysis of a reagent blank that all glassware and
reagent interferences are under control. Each time a set of sam-
ples is analyzed or there is a change in reagents, a laboratory
reagent blank should be processed as a safeguard against con-
tamination.
14.5 Procedural QC - The various steps of the analytical procedure
should have quality control measures. These include but are not
limited to:
14.5.1 GC performance - See Section 7.1 for performance cri-
teria.
14.5.2 MS performance - See Section 7.2 for performance cri-
teria.
D-38
-------
14.5.3 Qualitative identification - At least 10% of the PCB
identifications, as well as any questionable results,
should be confirmed by a second mass spectrometrist.
14.5.4 Quantitation - At least 10% of all manual calculations,
including peak area calculations, must be checked. After
changes in computer quantitation routines, the results
should be manually checked.
14.6 A minimum of 10% of all samples, one sample per month or one sam-
ple per matrix type, whichever is greater, selected at random,
must be run in triplicate to monitor the precision of the analy-
sis. An RSD of ±30% or less must be achieved. If the precision
is greater than ±30%, the analyst must be recertified (see Section
14.2).
14.7 A minimum of 10% of all samples, one sample per month or one sam-
ple per matrix type, whichever is greater, selected at random,
must be analyzed by the standard addition technique. Two aliquots
of the sample are analyzed, one "as is" and one spiked (surrogate
spiking and equilibration techniques are described in Section 9.2)
with a sufficient amount of Solution CSxxx to yield approximately
100 (jg/liter of each compound). The samples are analyzed together
and the quantitative results calculated. The recovery of the
spiked compounds (calculated by difference) must be 80-120%. If
the sample is known to contain specific PCB isomers, these isomers
may be substituted for solution CSxxx. If the concentrations of
PCBs are known to be high or low, the amount added should be ad-
justed so that the spiking level is 1.5 to 4 times the measured
PCB level in the unspiked sample.
14.8 Interlaboratory comparison - Interlaboratory comparison studies
are planned. Participation requirements, level of performance,
and the identity of the coordinating laboratory will be presented
in later revisions.
14.9 It is recommended that the participating laboratory adopt addi-
tional QC practices for use with this method. The specific prac-
tices that are most productive depend upon the needs of the lab-
oratory and the nature of the samples. Field duplicates or
triplicates may be analyzed to monitor the precision of the sam-
pling technique. Whenever possible, the laboratory should per-
form analysis of standard reference materials and participate in
relevant performance evaluation studies.
15.0 Quality Assurance
Each participating laboratory must develop a quality assurance plan ac-
cording to EPA guidelines.7 The quality assurance plan must be submitted
to the Agency for approval.
D-39
-------
16.0 Method Performance
The method performance is being evaluated. Limits of quantitation;
average intralaboratory recoveries, precision, and accuracy; and inter-
laboratory recoveries, precision, and accuracy will be presented.
17.0 Documentation and Records
Each laboratory is responsible for maintaining full records of the analy-
sis. Laboratory notebooks should be used for handwritten records. GC/MS
data must be archived on magnetic tape, disk, or a similar device. Hard
copy printouts may be kept in addition if desired. QC records should
be maintained separately from sample analysis records.
The documentation must describe completely how the analysis was performed.
Any variances from the protocol must be noted and fully described. Where
the protocol lists options (e.g., sample cleanup), the option used and
specifics (solvent volumes, digestion times, etc.) must be stated.
D-40
-------
REFERENCES
1. Environmental Protection Agency, Organochlorine Pesticides and PCBs—
Method 608," Fed. Reg., 44, 69501-69509 (December 3, 1979).
2. Environmental Protection Agency, "Base/Neutrals, Acids, and Pesticides--
Method 625," Fed. Reg., 44, 69540-69552 (December 3, 1979), and subse-
quent revisions.
3. "Methods 330.4 (Titrimetric, DPD-FAS) and 330.5 (Spectrophotometric, DPD)
for Chlorine, Total Residual," Methods for Chemical Analysis of Water and
Wastes, U.S. Environmental Protection Agency, Environmental Monitoring and
Support Laboratory, Cincinnati, Ohio, March 1979, EPA 600-4/79-020.
4. Erickson, M. D., and J. S. Stanley, "Methods of Analysis for Incidentally
Generated PCBs Literature Review and Preliminary Recommendations," Interim
Report No. 1, EPA Contract No. 68-01-5915, Task 51, 1982.
5. Bellar, T. A., and J. J. Lichtenberg, "The Determination of Polychlorinated
Biphenyls in Transformer Fluid and Waste Oils," Prepared for U.S. Environ-
mental Protection Agency (1981). EPA-600/4-81-045.
6. American Society for Testing and Materials, "Standard Method for Analysis
of Environmental Materials for Polychlorinated Biphenyls," pp. 877-885,
in Annual Book of ASTM Standards, Part 40, Philadelphia, Pennsylvania
(1980). ANSI/ASTM D 3304-77.
7. Quality Assurance Program Plan for the Office of Toxic Substances, Office
of Pesticides and Toxic Substances, U.S. Environmental Protection Agency,
Washington, D.C., October 1980.
D-41
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-560/5-82-006
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Analytical Methods for By-Product PCBs—Initial
Validation and Interim Protocols
5. REPORT DATE
October 11, 1982
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S) Mitchell D. Erickson, John S. Stanley, Gil
Radolovich, Kay Turman, Karin Bauer, Jon Onstot, Donna
Rose, and Margaret Wickham
8. PERFORMING ORGANIZATION REPORT NO.
MRI Project No. 4901-A51
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Midwest Research Institute
425 Volker Boulevard
Kansas City, MO 64110
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
EPA 68-01-5915, Task 51
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Toxic Substances, Field Studies Branch
TS-798
Washington, DC 20460
13. TYPE OF REPORT AND PERIOD COVERED
Interim 4, 4/24-8/31/82 .
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
The task manager is David P. Redford; the project officer is Frederick W. Kutz.
16. ABSTRACT
This document presents proposed analytical methods for analysis of by-product PCBs in
commercial products, product waste streams, wastewaters, and air. The analytical
method for commercial products and product waste streams consist of a flexible approach
for extraction and cleanup of particular matrices. The 13C-labeled PCB surrogates are
added as part of a strong quality assurance program to determine levels of recovery.
The wastewater method is based on EPA Methods 608 and 625 with revisions to include use
of the 13c-labeled PCB surrogates. The air method is a revision of a proposed EPA
method for the collection and analysis of PCBs in air and flue gas emissions. Capil-
lary or packed column gas chromatography/electron impact ionization mass spectrometry
is proposed as the primary instrumental method. Response factors and retention times
of 77 PCB congeners relative to tetrachlorobiphenyl-d6 are presented in addition to
statistical analysis to project validity of the data and extrapolation of relative
response factors to all 209 possible congeners. Preliminary studies using the ISC-
labeled surrogates to validate specific cleanup procedures and to analyze several com-
mercial products and product wastes indicate that the proposed analytical methods are
both feasible and practical.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI field/Group
Polychlorinated biphenyls
PCBs
Incidentally generated
Analytical protocols
Air
Wastewater
Commercial products
Commercial, waste
Capillary column
Electron impact
EIMS
Response factors
Relative retention
Surrogates
streams
gas chromatography
onization mass spectromet
relative response factor
times
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
243
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
------- |